KR20120085137A - Multiple linear regression-artificial neural network hybrid model predicting energesis of ideal gas of pure organic compound - Google Patents

Abstract

PURPOSE: A MLR(Multiple Linear Regression)-ANN(Artificial Neuron Network) mixed model, predicting generating energy of ideal gas of a pure organic compound, is provided to receive a molecular descriptor included in an optimum MLRM(Multiple Linear Regression Model) in order to form an ANN outputting a standard state absolute entropy, thereby improving prediction performance. CONSTITUTION: Optimum molecule descriptors are extracted for an experimental data set. Experimental data is separated into a training set and a test set. An optimum MLRM for the training set is explored. If test performance about the test set satisfies standards, the optimum MLRM is decided. The optimum MLRM obtains a generating energy value of ideal gas through the explored MLRM for the experimental data set.

Description

Multiple Linear Regression-Artificial Neural Network Hybrid Model Predicting Energesis of Ideal Gas of Pure Organic Compound

The present invention belongs to a field of physical chemistry called physical property prediction and relates to a method for predicting with high accuracy the generated energy of an ideal gas which is one of several physical properties of an organic compound.

Knowing the exact values of the various properties of an organic compound in detail can be used to determine the feasibility of the use of the substance, to design the synthesis and purification processes, and to establish methods and conditions for storage, transport, use, and disposal. It is an important issue both industrially and academically because it is critical to all decision-making processes throughout the process. The most accurate way of knowing the value of the property of interest of an organic compound of interest is also an experiment, but it is true that it is quite costly and time consuming in many aspects, including the preparation of purified samples and the construction of an environment for accurate measurement. Therefore, it may not be possible. Therefore, as an alternative, many researchers have long tried to predict the exact value of various physical properties of organic compounds. As such, the prediction of physical properties has a long history, and new prediction methods are constantly appearing. At present, several prediction models with different accuracy and application range coexist.

Several prediction models have been proposed to date for the energy of generation of ideal gas, which is the property of interest of the present invention. The energy of the ideal gas (Enthalpy of Formation for Ideal Gas at 298K) refers to the energy of the ideal gas generated when the pure material is at 298.15K (absolute temperature). The results of previous studies on the prediction of the formation energy of an ideal gas are described in Polling BE, Prausnitz JM, O'Connell JP The Properties of Gases and Liquids , (5 ed.). New York: McGraw Hill. (2000).]. Currently known and widely used models for predicting the energy generated by ideal gases are those that use group contribution methods and quantum mechanical calculations. Table 1 shows, in chronological order, the major group contribution models for predicting the energy generation of ideal gases that have been proposed.

year proponent 1968 Benson 1969 Benson & Buss 1971 Thinh, T.-P. & T. K. Trong 1984 Joback 1992 Ratkey & harrison 1994 Constantinou & Gani 1994 Mavrovouniotis & Constantinou 1997 Forsythe & Karp 1997 Steele

A typical form of group contribution model is given by

를 구하기 위해서는 먼저 값을 알고자 하는 화합물의 분자를 미리 정해진 다수의 조각형식들에 맞추어 쪼갠 다음 각 조각형식들의 개수

를 구한다. 이를 다시 그 형식에 할당된 계수

와 곱한 것을 합산한 것이 예측값

가 된다. 계수

들은 실험값이 존재하는 화합물들로부터 모형이 최선의 성능을 갖도록 통계적인 방법을 통해 결정된다.Property value

In order to find the value, first divide the molecule of the compound whose value you want to know according to a number of pieces.

. This is again the coefficient assigned to that type

Multiplying and multiplied by

Becomes Coefficient

These are determined by statistical methods to ensure that the model has the best performance from the compounds with experimental values.

These group contribution methods have been somewhat successful in the past, but lack the theoretical basis, and sometimes the only way to split them into pieces is not, or even nonexistent, which makes the calculation of values impossible. Also, as the model is improved to improve the predictive performance, it becomes more complicated and difficult to handle.

One of the alternatives to the group contribution method in constructing the prediction model is the QSPR (quantitative structure-property relationship) method. This method basically starts with the assumption that the properties of a compound are a function of the structural properties of the molecule and uses various molecular descriptors that reflect different structural properties. The number of molecular descriptors proposed to date has been thousands, and many kinds of molecular descriptors range from simple ones such as the number of carbons or hydrogens in a molecule to complex ones such as the shape, connection state, and electrochemical properties of molecules. Calculation methods have been developed [Todeschini R., V. Consonni V., Molecular Descriptors for Chemoinformatics: Second, Revised and Enlarged Edition: Volume I / II , Wiley-VCH, 2009]. The QSPR prediction model is presented in the form of a function that includes some of these molecular descriptors and sometimes, in addition, some of the other physicochemical properties of the compound, which are also functions of structural properties.

One of the alternatives to the group contribution method in constructing the prediction model is the quantum mechanical calculation method. However, the quantum mechanical calculation method has a big disadvantage in that it takes too much calculation time instead of calculating an accurate prediction value when a high dimensional calculation is performed. Therefore, after the appropriate level of quantum mechanical calculation, we selected a method to correct the error of the predicted value caused by the low level of quantum mechanical calculation. As a method of correcting the error of quantum mechanical calculation, a quantitative structure-property relationship (QSPR) method was used. This method basically starts with the assumption that the properties of a compound are a function of the structural properties of the molecule and uses various molecular descriptors that reflect different structural properties. The number of molecular descriptors proposed to date has been thousands, and many kinds of molecular descriptors range from simple ones such as the number of carbons or hydrogens in a molecule to complex ones such as the shape, connection state, and electrochemical properties of molecules. Calculation methods have been developed [Todeschini R., V. Consonni V., Molecular Descriptors for Chemoinformatics: Second, Revised and Enlarged Edition: Volume I / II , Wiley-VCH, 2009]. The QSPR prediction model is a function of functions that include some of these molecular descriptors, and sometimes, in addition, some of the compound's AOMG (Atomic Orbital Molecular Graph) molecular markers in their atoms, which are also a function of structural properties. Presented in form.

들의 선형 결합 함수이며 각 계수

들은 주로 다중선형회귀분석을 통해 실험데이터로부터 결정된다.The most frequently adopted form of such a function is

Linear coupling function of each coefficient

These are mainly determined from experimental data through multiple linear regression analysis.

The use of hybrid orbital species as atoms in the atom is a modification of the idea of AOMG (Atomic Orbital Molecular Graph), and the energy contributing to the energy generated by the ideal gas depends on the type of hybrid orbital by atom. I applied the idea that it would be different. A detailed description of the AOMG pattern is given later in this document in the detailed description of the invention.

이 부여되어 있다. 은닉층과 출력층의 각 노드들은 전 단계의 노드들로부터 이러한 연결들을 통해 입력을 받은 뒤 이를 가공하여 출력값을 만드는데 이때 활성화 함수(activation function)라 불리는 함수

를 적용한다. 이러한 인공신경망을 실제로 활용하려면 먼저, 다양한 입력값과 그 입력값에 대응하는 출력값을 함께 묶어 놓은 샘플집합을 이용하여 인공신경망을 훈련시키는 과정이 필요한데 이는 주어진 입력에 대한 인공신경망의 출력과 원하는 출력의 차이가 최소가 되도록 역전파(back propagation) 알고리즘을 사용하여 각 연결의 가중치를 최적화 하는 것을 말한다. 이러한 훈련을 거친 인공신경망은 문제해결에 필요한 규칙이나 지식을 따로 제공하지 않아도 학습을 통해서 스스로 일반적인 규칙을 수립하여 미지의 입력에 대해서도 타당성 있는 출력을 내주므로 화합물의 물성예측과 같이 기반 이론이 결여되어 있는 분야에 매우 유용한 수단으로 널리 이용되고 있다.Another way to create a QSPR model is to use an artificial neural network. Artificial neural network technology is one of the long-standing research results of human beings who want to make intelligent machines by modeling human nerve cells with intelligence, and is an information processing technology that has been applied in various fields since the mid 20th century. . 2 shows a typical example of an artificial neural network. As can be seen, the artificial neural network has an input layer for receiving input data, an output layer for producing output data, and a hidden layer located between them, each layer having one or more nodes. ) Each node in the hidden layer is connected to nodes in the input and output layers, and each connection has a quantity called a weight.

Is granted. Each node in the hidden and output layers receives input through these connections from the nodes in the previous stage and then processes it to produce an output, which is called an activation function.

Apply. In order to actually use such an artificial neural network, first, the neural network is trained using a sample set in which various input values and output values corresponding to the input values are bundled together. This means optimizing the weight of each connection using a back propagation algorithm to minimize the difference. The neural network that has undergone such training does not provide the necessary rules or knowledge to solve the problem, but it establishes general rules through learning and gives valid output for unknown inputs. It is widely used as a very useful means in the field.

The QSPR prediction model using AOMG molecular descriptors for the generation energy of an ideal gas is not widely proposed, and the proposed ones also have disadvantages in that the number of sample compounds is not large or the application target is limited to a specific series of compounds. Andres Mercader, Eduardo A. Castro, Andrey A. Toropov QSPR modeling of the enthalpy of formation from elements by means of correlation weighting of local invariants of atomic orbital molecular graphs. Chemical Physics Letters 330 (2000) 612-623 reported a model with a coefficient of determination of 0.99102 using data for 65 compounds.

The technical problem to be solved by the present invention is to overcome the shortcomings of the existing models mentioned above and to show more broad and more accurate prediction performance, hydrogen (H), carbon (C), nitrogen (N), oxygen (O), sulfur ( It is to construct a multilinear regression-artificial neural network hybrid model for the energy of generation of ideal gas of a pure organic compound composed of less than 5 elements including S) and molecules having 25 or less atoms except hydrogen.

We have achieved this goal by constructing a multilinear regression model that considers a wider variety of molecular presenters based on more experimental data on the basis of the quantum mechanical calculations of the energy of formation of ideal gases. The above-mentioned limitations on the range of organic compounds to which the predictive model can be applied are mainly due to the current state of the art when some of the molecular descriptors used require quantum mechanical calculations. This is due to the fact that for the compounds beyond the stated range, problems arise in terms of accuracy and calculation time. However, even within the above limitations, since there are a great many compounds and industrially important compounds are included in a large amount, it is determined that the present invention can greatly benefit human society.

Humans today depend on a wide variety of organic compounds, including plastics, fibers, rubber, paints, fertilizers, medicines and fuels, and this trend is expected to intensify. According to the American Chemical Society (ACS), as of July 2010, the total number of compounds registered was over 54,000,000. In comparison, even if the physical property value is only one, the number of experimentally known compounds is only tens of thousands. The physical property value of compounds is essential for the better life of mankind, such as the development of new materials and new drugs, the optimal design of chemical plants, the improvement of the productivity of existing facilities, the development and saving of resources, the securing of safety, and the protection of the environment. In particular, the energy generated by the ideal gas is an important property that commercial programs such as AspenPlus and Pro / II, which are well known as the optimum design programs for chemical plants, are urgently requesting the exact values. However, at present, the number of compounds whose experimental values are known is only a few thousand, and depending on the compounds, the work of obtaining data through experiments may be past due to toxicity, instability, and difficulty of purification. From this point of view, the present invention, which provides a high accuracy of generating energy of an ideal gas of a large number of compounds with only information on a molecule without undergoing experiments, not only saves cost and time, but also estimates the value even when the experiment is impossible. In addition to facilitating the R & D activities of related industries, it also has the effect of providing the appropriate information to academics and academia, wherever they need it, to facilitate its activities. would.

1 is a flowchart illustrating a process of constructing a multiple linear regression-artificial neural network hybrid model for energy generated by an ideal gas provided by the present invention.
FIG. 2 shows AOMG types that a pure compound used in the multiple linear regression-artificial neural network hybrid model for generating energy of an ideal gas provided by the present invention.
Figure 3 shows the covalent bonds of AOMG types that a pure compound used in the multiple linear regression-artificial neural network hybrid model for the energy generated by the ideal gas provided by the present invention.
Figure 4 is a view showing the structure of the artificial neural network used in the present invention.
FIG. 5 is a parity diagram of 1407 experimental data of a Joback model which is a group contribution model of generated energy of an ideal gas.
FIG. 6 is a parity diagram of 1318 experimental data of a Gani model, which is a group contribution model for generating energy of an ideal gas.
FIG. 7 is a parity diagram of 1536 experimental data of a multiple linear regression-artificial neural network hybrid model for energy generated by an ideal gas provided by the present invention.
8 is a histogram diagram of 1407 experimental data of the Joback model.
9 is a histogram diagram of 1318 experimental data of the Gani model.
FIG. 10 is a histogram plot of 1536 experimental data of a multiple linear regression-artificial neural network hybrid model.

Figure 1 is a simplified representation of the process of building a multiple linear regression-artificial neural network hybrid model for the energy generated by the ideal gas.

The first thing to do when building the model is to collect and review the experimental data as specified in step 1. As a result of extensive research on all literatures and data that can be referred to through various papers, books, Internet sites, etc. for the present invention, the data of the energy of generation of ideal gas for the compounds meeting the conditions of the present invention in total 4887. Was collected. The collected data were examined in a variety of ways to determine whether they were truly valid values that could be used to build the model.They were not experimental values, errors in data notation, or values for the same compound. In the case of unreliably distant values, data were analyzed in detail, and data were modified or deleted to finally select data for a total of 2041 compounds. In addition, when constructing the property prediction model, it was possible to classify the sample compounds into pure molecules and molecules including radicals, and to model them separately, in light of the previous experience in terms of predictive performance. In the light of the previous experience, it was better to model separately by classifying carbon and hydrogen-only hydrocarbons and non-hydrocarbons separately. So they decided to model them by classifying them into molecules containing 524 radicals, 663 hydrocarbons, and 854 non-hydrocarbons. In addition, two types of molecules containing radical centers and those with interactions between resonance structures and radical centers and simple modeling of molecules containing radical centers could be used to improve the accuracy of prediction. . In the present invention, the 'organic compound' or 'compound' is composed of five elements such as hydrogen (H), carbon (C), nitrogen (N), oxygen (O), and sulfur (S), and atoms except hydrogen Refers to a substance consisting of molecules of 25 or less.

The next step is to prepare predictions with quantum mechanical calculations for these compounds. In order to calculate the electronic structure of molecules, the Schrodinger equation is usually solved by a pure theory to solve the electron energy. However, in the case of systems with many electrons, Hartley uses an approximation method that ignores electron correlation. Hartree-Fock (HF) method [CC J. Roothan, Rev. Mod. Phys. 23, 69 (1951)]. This approximation introduces a fundamental error in the calculated results and adds a multidimensional theoretical perturbation term to the Post Hartree-Fock method [C. Moller and M. S. Plesset, Phys. Rev. 46, 618 (1934)], to obtain a more accurate solution, but require a relatively large amount of computation. In this way, it is too costly or time-consuming to calculate large molecules.

In addition, the Gaussian method that combines Hartley-Fork and Post Hartley-Fork [L. A. Curtiss, K. Raghavachari, G. W. Trucks, and J. A. Pople, J. Chem. Phys. 94, 7221 (1991); L. A. Curtiss, K. Raghavachari, P. C. Redfern, V. Rassolov, and J. A. Pople, J. Chem. Phys. 109, 7764 (1998) show very little error in energy prediction, but more computation is required because it performs energy calculations for several post-Hartley-Fork methods.

Density functional theory is used to find the ground state using the function of the total energy using the electron density function instead of the wave function with the multidimensional perturbation term to consider the correlation between the electrons of the molecules of many electrons. R. Seeger and JA Pople, J. Chem. Phys. 66, 3045 (1977). The advantage of the density functional theory is that the electron density only needs to be taken into account so that more accurate results can be obtained with comparable calculations to the Hartree-Fock method. The combination of exchange functions and correlation functions for calculating the exchange-correlation energy of the electrons is used to obtain more improved results without increasing the calculation amount.

The computational theories previously attempted to select an optimal quantum mechanical calculation method are the density ranges of the aforementioned Hartley-Fork method, various Post-Hartley-Fork methods, Gaussian (G2, G3) methods, and various combinations of functional functions. Function theory. Among them, one method of density functional theory, which is the best performance calculation time, was selected.

Therefore, in the present invention, the optimization of the molecular structure and the frequency calculation are performed by applying the calculation method of the specified density functional theory using a commercial quantum mechanical calculation program.

The next step is to prepare AOMG values to correct for errors in quantum mechanical calculations. First, I will explain the kind of hybrid orbitals of compounds composed of pure molecules. Since the target molecules we are dealing with consist only of C, H, N, O, and S, the combination of constituent atoms and bond forms of a general organic compound is as shown in Figs. 3 are patterns included in covalent bonds of carbon, nitrogen, and oxygen atoms in a molecule. The presence of covalent bonds results in a special stabilization of the overall energy of the molecule, which is also a separation pattern in which each atom in the molecule contributes to energy. Thus, 17 AOMG patterns can be defined for all molecules consisting of C, H, N, O, and S only. These 17 AOMG patterns apply to all pure molecular compounds, and molecular compounds with radical centers apply 15 AOMG patterns, including only the C5 pattern, among the AOMG patterns that can appear in the covalent bonds shown in FIG. As a result of the predictive performance thus far, pure molecular hydrocarbons have been shown to perform fairly well, and no further process has been performed. And the predictive performance of non-hydrocarbons, molecules with no interaction between radical center and resonance structure, was not satisfactory, but showed good predictive performance. Therefore, rather than adding molecular markers, the neural network was used to increase the predictive performance by considering the nonlinearity between molecular markers and predictive properties.

The next step is to prepare the values of molecular descriptors for the compounds that have resonance structures and interactions with the radical centers. This document prepares the values of the molecular presenters, which are collectively referred to as the quantum mechanical calculation and the calculation of the values for the AOMG pattern. The values for various molecular descriptors totaling up to 1978 were collectively calculated by computer from files containing information about the molecules of each compound, which was then unsuitable, i.e. for all sample compounds. These same ones are selected for those that cannot be independent variables of the model. This is because reducing the number of molecular descriptors can reduce the computation time required to find the optimal model.

In step 4, the sample compounds are divided into two parts: the training set used to explore the predictive model, and the test set used to test the predicted performance of the determined model. The set of sample compounds to be used in this model is divided roughly into pure hydrogenated molecules with radical centers and without radical centers. And molecules with radical centers are divided into structures with and without resonance centers. Each of the three sample compound sets were divided into 5: 5 to 8: 2, preferably 6 to 4, training groups and test sets, taking care not to distribute similar molecules on one side.

이며 이들을 다 조사하는 것은 현실적으로 불가능하다.The genetic algorithm is then based on a training set for molecules with interactions between radical centers and resonance structures in the training set [Judson, "Genetic Algorithms and Their Uses in Chemistry", Reviews in Computational Chemistry, Lipkowitz & Boyd. , Eds., Vol. 10, pp.1-73 (VCH Publishers, NY, 1997)] find the best multiple linear regression model. The term 'best' is used in the sense of relative meaning that it can be obtained in a relatively short time and has a performance very close to the optimal solution in the absolute sense. The reason for not finding the optimal solution directly is because of the long computation time. For example, when there are 1700 suitable molecular representations out of 1978 molecular representations, you can choose from five different linear regression models. The total number is

It is practically impossible to investigate them all.

In order to obtain useful results within a limited time, the present invention employs a genetic algorithm, and the detailed method is as follows. First, a population of multiple linear regression models is created by randomly drawing a certain number of molecular descriptors from a pool of molecular descriptors. For example, suppose you created a population of 1000 different polylinear regression models that were randomly drawn from five of the 1700 suitable molecular descriptors.

Individuals, called chromosomes, are coded by combining the numbers of extracted molecular descriptors. For example, the chromosome of the multiple linear regression model formed by the 45th, 167, 684, 1033, and 1502th molecular descriptors among 1700 molecular descriptors can be expressed as (45, 167, 684, 1033, 1502). Two parent chromosomes are selected from the populations thus generated and crossed over to generate children. In the present invention, the Roulette Wheel method is adopted as a selection method for selecting the parent chromosomes.

) 또는 평균절대오차(average absolute error: AAE)를 활용하였다. 즉 결정계수값이 크거나 평균절대오차값이 작은 것이 선택확률이 높도록 하였다.The Roulette Wheel method is the most commonly used selection algorithm, which allocates a roulette region to the chromosome in proportion to the fitness of each chromosome, and then rotates the roulette to select the chromosome of the corresponding region. Therefore, in this method, the higher the fit, the more likely it is to be selected. The coefficient of determination of the regression model is used to calculate the goodness of fit for each chromosome that determines the probability of selection.

) Or average absolute error (AAE). In other words, the larger the coefficient of determination or the smaller the mean absolute error, the higher the probability of selection.

The single point crossover method is adopted as the breeding method. The most common breeding method is to generate a child by selecting one crossing point on the parent chromosome and exchanging chromosomal parts before and after the point. For example, if the parent chromosome is given as (24, 262, 343, 789, 1290), (38, 454, 554, 1322, 1449), and there is a crossing point between the third and fourth elements, then the child chromosomes are (24, 262, 343, 1322, 1449), (38, 454, 554, 789, 1290).

When the offspring are created, they have a chance of mutating a portion of their chromosomes, which randomly replaces several elements with completely new values. This new information does not exist in the current population. It is a process that makes it possible to search to compensate for the case where there is no proper solution in the initial set combination.

In this way, a new generation of populations are created by replacing some or all of the existing populations with newly obtained entities. Repeat this process until the number of generations reaches a pre-determined value (usually between 10 and 1000), and then select and end up the regression model with the best predictive performance.

Once this best multiple linear regression model has been selected, the next step is to examine its validity. If a problem is found that the t-test value of the molecular descriptors included in the model is not good, go back and look for another model. For example, if the number of sample compounds is 1005 and the selected model consists of five molecular descriptors, if the t-test for one of the molecular descriptors is 3.3 or higher, then the probability that the molecular descriptor is irrelevant to that property is 0.1 It means less than%. In the present invention, when there is a molecular presenter having a t-test value of less than about 3, the selected model is discarded and another model is found. In addition, even if the values of the molecular descriptors for the sample compounds are the same except for a few few compounds, they are not considered to be reliable models. In general, increasing the number of molecular expressions included in the model increases the predictive performance, but such problems occur. Therefore, the final model is usually obtained by repeating these steps through several trials and errors while changing the number of molecular expressions. If the problem no longer appears in the selected model, proceed to the next step.

Next, in Step 7, assess the predictive performance of the found model using test sets that did not participate in model formation. If a problem is found in the training set that results in much lower predictive performance or samples that are significantly off predicted, go to step 4 and readjust the training and test sets before proceeding. If the difference between the training set and the test set does not exceed 20% of the absolute mean error (AAE) obtained for the training set, it is judged that the predictive performance is satisfied.

In this way, a neural network can be established once a multiple linear regression model of three sample compound sets for standard production energy (groups with resonance structures and radical center interactions, groups without resonance structures and radical center interactions, and non-hydrocarbon groups) is established. In order to establish the model, we first standardize the data of the molecular presenters and the experimental data of the standard generation energy, that is, subtract the mean of the data from each value and divide by the standard deviation. The entire sample thus prepared is divided into a training set, a validation set, and a test set in an approximately 6: 2: 2 ratio.

을, 출력층의 활성화 함수로는 선형함수 즉

를 채택하였다. 따라서 입력층의 각 노드들이 받는 입력값들을

라 할 때 은닉층의 j번째 노드의 출력값은

와 같이 주어지며 은닉층이

개의 노드로 이루어져 있을 때 출력층 출력노드의 최종 출력값은

와 같이 주어진다. 여기서

는 문턱 가중치(threshold weight)를 의미한다.We then use them to find the best artificial neural network model. In this case, the search range is limited to a neural network having a hidden layer between the input layer and the output layer as shown in FIG. 2 and having three structures connected only in a feed forward direction, that is, in a direction from the input to the output. It was. The input layer is composed of the same number of nodes that receive the values of the molecular markers included in the already established multiple linear regression model, and the output layer is composed of one node that outputs the critical volume. In addition, as the activation function of the hidden layer, the Sigmoid function,

, The activation function of the output layer is a linear function

Was adopted. Therefore, the input values that each node in the input layer receives

In this case, the output value of the j th node of the hidden layer is

Is given by

The final output value of the output layer output node when composed of four nodes

Is given by here

Denotes a threshold weight.

들의 다양한 초기값세트(보통 1000세트이내)를 마련하고, 훈련집합을 사용하여 각 세트로 초기화된 신경망을 역전파 알고리즘을 통해 반복 훈련함으로써 가중치

들의 최적화된 값을 찾는다. 최적화에 대한 판단은 매 훈련 후 경신된 가중치들의 값으로 정해지는 모형을 검증집합에 적용하였을 때 그 평균제곱오차(mean square error)의 값이 최소가 되는 것으로 한다. 보통은 3000~5000번의 반복훈련 내에 이러한 시점이 나오게 된다. 이렇게 얻어진 각 초기값세트에 대응하는 최적화된 신경망모형을 훈련집합, 검증집합, 시험집합에 각각 적용하여 그 평균제곱오차들이 모두 다중선형회귀모형의 그것들보다 작은 것만을 모은다. 이러한 것이 여러 개 있을 경우, 결정계수나 평균절대오차 등을 기준으로 가장 우수한 모형을 선택한다.The search proceeds from increasing the number of hidden nodes to one in order. Usually, the search proceeds to twice the number of input nodes. However, if a satisfactory model is not found, the search proceeds further. The detailed procedure is as follows. First, the weight generated by using random number generation function for each number of hidden nodes

Prepare different sets of initial values (usually within 1000 sets) and weight them by repeatedly training the neural networks initialized with each set using the back-up algorithm.

Find the optimal value for these. The judgment of the optimization is that the mean square error is minimized when the model, which is determined by the updated weights after each training, is applied to the test set. Normally this will occur within 3000 to 5000 repetitions. The optimized neural network model corresponding to each set of initial values thus obtained is applied to the training set, the test set, and the test set, respectively, to collect only those whose mean square errors are smaller than those of the multiple linear regression model. If there are several of these, choose the best model based on the coefficient of determination or the absolute absolute error.

When the artificial neural network model is selected, an overfitting prevention standard is finally set. This is a measure to improve the instability that the neural network gives wrong answers to unknown inputs as a result of excessive training.If the absolute value of the difference between the predicted values of the neural network model and the multiple linear regression model exceeds the reference value by setting a reference value, It means to adopt the predictive value of the multiple linear regression model and to adopt the value of the artificial neural network model if it is smaller than this.

Through this process, the results of establishing the multiple linear regression-artificial neural network hybrid model for the generation energy of the ideal gas are shown in Table 2 (the main contents of the QSPR prediction model for the generation energy of the hydrocarbon ideal gas) and Table 3 (non-hydrocarbon). The main contents of the QSPR prediction model for the generation energy of the ideal gas in Table 4 (the main contents of the QSPR prediction model for the generation energy of the ideal gas of the molecule having no interaction between the resonance structure and the radical center) and Table 5 The main content of the QSPR predictive model for the generation energy of an ideal gas of a molecule having an interaction between a resonance structure and a radical center is briefly described.

Number of sample compounds 663 Number of molecular descriptors 6 Names of Molecular Presenters P ₁ : Energy of Formation for Ideal Gas at 298K from quantum mechanics
P ₂ : Atomic Orbital Molecular Graph Pattern -H1
P ₃ : Atomic Orbital Molecular Graph Pattern -C2
P ₄ : Atomic Orbital Molecular Graph Pattern -C3
P ₅ : Atomic Orbital Molecular Graph Pattern -C4
P ₆ : Atomic Orbital Molecular Graph Pattern -C5 Coefficient of determination 0.998878 Standard error 1.250625 kcal / mol Mean Absolute Error 0.758519 kcal / mol Generated energy of model ideal gas

Generated energy of ideal gas

Number of sample compounds 854 Number of molecular descriptors 12 Names of Molecular Presenters

: 양자역학 이상기체의 생성 에너지(Enthalpy of Formation for Ideal Gas at 298K from quantum mechanics)

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴- N1

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴- N2

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴- N3

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴- N4

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴- O1

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph)패턴- O2

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴- O3

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴- S1

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴- S2

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴- S3

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴- S4

: Enthalpy of Formation for Ideal Gas at 298K from quantum mechanics

: Atomic Orbital Molecular Graph Pattern-N1

: Atomic Orbital Molecular Graph Pattern-N2

: Atomic Orbital Molecular Graph Pattern-N3

: Atomic Orbital Molecular Graph Pattern-N4

: Atomic Orbital Molecular Graph Pattern-O1

: Atomic Orbital Molecular Graph Pattern-O2

: Atomic Orbital Molecular Graph Pattern-O3

: Atomic Orbital Molecular Graph Pattern-S1

: Atomic Orbital Molecular Graph Pattern-S2

: Atomic Orbital Molecular Graph Pattern-S3

: Atomic Orbital Molecular Graph Pattern-S4
Regression Model Decision Coefficients 0.997871 Regression Model Mean Absolute Error 1.85253

Regression model Energy generated by an ideal gas [

] =

Artificial Neural Network Determination Coefficient 0.997871 Artificial neural network mean absolute error 0.827643

Artificial Neural Network Model Generated energy of ideal gas
=

Overconformity Prevention Criteria 20

Number of sample compounds 358 Number of molecular descriptors 15 Names of Molecular Presenters

: 양자역학 이상기체의 생성 에너지(Enthalpy of Formation for Ideal Gas at 298K from quantum mechanics)

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -H1

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -C2

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -C3

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -C4

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -C5

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -N1

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -N2

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -N3

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -O1

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -O2

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -S1

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -S2

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -S3

: 원자 궤도함수 분자 그래프(Atomic Orbital Molecular Graph) 패턴 -S4

: Enthalpy of Formation for Ideal Gas at 298K from quantum mechanics

: Atomic Orbital Molecular Graph Pattern -H1

: Atomic Orbital Molecular Graph Pattern -C2

: Atomic Orbital Molecular Graph Pattern -C3

: Atomic Orbital Molecular Graph Pattern -C4

: Atomic Orbital Molecular Graph Pattern -C5

: Atomic Orbital Molecular Graph Pattern -N1

: Atomic Orbital Molecular Graph Pattern -N2

: Atomic Orbital Molecular Graph Pattern -N3

: Atomic Orbital Molecular Graph Pattern -O1

: Atomic Orbital Molecular Graph Pattern -O2

: Atomic Orbital Molecular Graph Pattern -S1

: Atomic Orbital Molecular Graph Pattern -S2

: Atomic Orbital Molecular Graph Pattern -S3

: Atomic Orbital Molecular Graph Pattern -S4 Regression Model Decision Coefficients 0.991293 Regression Model Mean Absolute Error 1.730095

Regression model Energy generated by an ideal gas [

] =

Artificial Neural Network Determination Coefficient 0.995293 Artificial neural network mean absolute error 1.016031

Artificial Neural Network Model Generated energy of ideal gas
=

Overconformity Prevention Criteria 20

Number of sample compounds 524 Number of molecular descriptors 16 Names of Molecular Presenters P ₁ : Energy of Formation for Ideal Gas at 298K from quantum mechanics
P ₂ : Atomic Orbital Molecular Graph Pattern-C2
P ₃ : Atomic Orbital Molecular Graph Pattern-C3
P ₄ : Atomic Orbital Molecular Graph Pattern-C4
P ₅ : Atomic Orbital Molecular Graph Pattern-C5
P ₆ : Atomic Orbital Molecular Graph Pattern -N2
P ₇ : Atomic Orbital Molecular Graph Pattern -N3
P ₈ : Atomic Orbital Molecular Graph Pattern -O1
P ₉ : Atomic Orbital Molecular Graph Pattern -O2
P ₁₀ : Atomic Orbital Molecular Graph Pattern-S1
P ₁₁ : Atomic Orbital Molecular Graph Pattern -S2
P ₁₂ : Atomic Orbital Molecular Graph Pattern-S3
P ₁₃ : Atomic Orbital Molecular Graph Pattern-S4
P ₁₄ : Balaban Index
P ₁₅ : Area weighted surface charge fraction of hydrogen bond dependent hydrogen bond base atoms (HA dependent HDCA-2 / TMSA)
P16: Kier flexibility index
Regression Model Decision Coefficients 0.993733 Mean Absolute Error

1.350155

Regression model Energy generated by an ideal gas [

] =

의 평균절대오차값을 가짐을 알게 되었다. 또한 Gani 모형은 1318개에 대해서만 예측값을 계산해주며 0.991044의 결정계수값과 2.625765

의 평균절대오차값을 가짐을 알게 되었다. 반면 다중선형회귀-인공신경망 혼성모형은 1536개 전부에 대해 예측값을 계산해주며 0.996184의 결정계수값과 1.892156

의 평균절대오차값을 가져 다른 두 모형보다 우수함을 알게 되었다. 도 5, 6, 7는 각 모형의 예측성능을 보여주는 패리티(parity) 도면들이며 이 도면들로부터 다중선형회귀-인공신경망 혼성모형이 다른 두 모형보다 우수한 성능을 가짐을 눈으로 확인할 수 있다. 한편 1536개 화합물들에 대한 실험데이터 중 실험오차가 알려진 것들의 평균 오차는 약 3

이며 이 값을 중심으로 실험값과 예측값 사이의 오차를 히스토그램으로 그린 것이 도 8, 9, 10이다. 이 도면들은 Joback 모형은 66.09%, Gani 모형은 76.66%, 다중선형회귀-인공신경망 혼성모형은 86.25%의 확률로 평균 실험오차의 범위 이내로 이상기체의 생성 에너지값을 예측하고 있음을 보여주어 다중선형회귀-인공신경망 혼성모형이 다른 두 모형보다 정확함을 증명해준다.In order to show that the present invention is superior to the existing technology, the multilinear regression-artificial neural network hybrid model thus established and the existing group contribution model widely used, namely Joback [Joback KG, Reid RC, Estimation of pure-component properties from group-contributions, Chem. Eng. Comm. , 57: 233 (1987).] And the Gani model [Constantinou, L., Gani R., New Group Contribution Method for Estimating Properties of Pure Compounds, AIChE ., 40: 1697 (1994).]. Comparisons were made using data of 1535 known compounds. As a result, the Joback model calculates the predictions for only 1407, and the coefficient of determination of 0.985271 and 8.022368

It is found that the mean absolute error of. In addition, the Gani model calculates predictions for only 1318, with a coefficient of determination of 0.991044 and 2.625765.

It is found that the mean absolute error of. On the other hand, the multiple linear regression-artificial neural network hybrid model calculates the predicted values for all 1536, and has a coefficient of determination of 0.996184 and 1.892156.

The average absolute error of is better than the other two models. 5, 6, and 7 are parity diagrams showing the predictive performance of each model, and it can be seen from these figures that the multiple linear regression-artificial neural network hybrid model has better performance than the other two models. Meanwhile, the average error of the experimental data of 1536 compounds with known experimental errors was about 3

8, 9, and 10 show a histogram of the error between the experimental value and the predicted value. These figures show that the Joback model is 66.09%, the Gani model is 76.66%, and the multiple linear regression-artificial neural network hybrid model has a probability of 86.25%. The regression-artificial neural network hybrid model proves to be more accurate than the other two models.

The present invention is not limited to the above-described embodiments, and any person having ordinary skill in the art to which the present invention pertains may make various modifications without departing from the gist of the present invention as claimed in the claims. Such changes are intended to fall within the scope of the claims.

Claims (50)

A first step of inputting hydrocarbon-based experimental data into experimental data of collected sample organic compounds;
A second step of preparing a molecular presenter value for the energy of generation of the ideal gas of the hydrocarbon compound through the experimental data input in the first step;
A third step of extracting optimal molecular presenters for each experimental data set input in the first step;
A fourth step of separating the experimental data input in the first step into a training set and a test set;
A fifth step of searching for an optimal multiple linear regression model for the training set;
A sixth step of examining the validity of the selected model;
A seventh step of repeating the fifth and sixth steps if there is no validity in the sixth step, and testing the predictive performance of the model with respect to the test set if it is valid;
An eighth step of repeating steps 4 to 7 if the performance does not satisfy the criterion in the seventh step test for the test set, and determining an optimal multilinear regression model if the criterion is satisfied; And
A ninth step of obtaining a generated energy value of an ideal gas through the multilinear regression model searched for the experimental data set of the collected sample organic compounds by the optimal multilinear regression model satisfying the performance test in the eighth step A method for obtaining the energy of generation of an ideal gas of a hydrocarbon organic compound through a multiple linear regression model comprising a step.
The energy generation of an ideal gas of a hydrocarbon organic compound according to claim 1, wherein the optimal molecular descriptor in the third step is an independent molecular descriptor whose values are not the same for all sample compounds. How to obtain.
The method of claim 1, wherein the training set and the test set are divided by a ratio of 5: 5 to 8: 2 in the fourth step.
The method according to claim 1, wherein in the fifth step, the multiple linear regression model searches for a multiple linear regression model by applying a genetic algorithm to the training set. A method for obtaining the energy of generation of ideal gas of a compound.
The method of claim 4, wherein the genetic algorithm generates a population composed of a plurality of multiple linear regression models randomly drawn from a predetermined number of molecular representations in a pool of molecular representations. Making; Encoding each individual by combining the numbers of the extracted molecular presenters; Selecting two parent chromosomes from the created population by the Roulette Wheel method and generating offspring by a single point crossover method; Ideal gas of the hydrocarbon organic compound through a multi-linear regression model comprising the step of mutating a portion of the chromosome of the generated offspring with a certain probability and then replacing a part of the existing population with them to create a new population How to find the generated energy of.
The method of claim 1, wherein the fifth step includes determining the predictive performance by determining the predictive performance based on the crystal coefficient or the mean absolute error of the regression model.
The method of claim 1, wherein the validity in the sixth step is to determine the generated energy of the ideal gas of the hydrocarbon organic compound through a multiple linear regression model in which the validity is determined by the t-test value.
The method of claim 1, wherein in the eighth step, if the predictive performance of the test set is similar to the predicted performance of the training set, a multiple linear regression model is determined, and the predictive performance of the test set is different from the predicted performance of the training set. After that, a multilinear regression model that reclassifies the training set and the test set is used to obtain the generated energy of the standard gas ideal gas of the hydrocarbon organic compound.
A first step of inputting non-hydrocarbon-based experimental data among collected sample organic compounds;
A second step of preparing a molecular presenter value for energy generated by an ideal gas of the non-hydrocarbon-based organic compound of sample organic compounds;
Extracting optimal molecular descriptors;
A fourth step of separating the experimental data into a training set and a test set;
A fifth step of searching for an optimal multiple linear regression model for the training set;
A sixth step of examining the validity of the selected model;
A seventh step of repeating the fifth and sixth steps if there is no validity in the sixth step, and testing the predictive performance of the model with respect to the test set if it is valid;
An eighth step of repeating steps 4 to 7 if the performance does not satisfy the criterion in the seventh step test for the test set, and separating the three sets after sample normalization if the criterion is satisfied;
A ninth step of searching for an optimal neural network model after dividing the entire sample into three sets;
The absolute value of the difference between the predicted energy produced by the ideal linear regression model satisfying the performance test in the eighth step and the predicted energy produced by the ideal gas calculated by the optimal neural network model found in the ninth step. A tenth step of comparing with a preset overfit prevention reference value; And
If the difference is larger than the reference value for preventing overfitting, the generated energy predicted value of the ideal gas obtained by the multiple linear regression model obtained in the eighth step is adopted as the generated energy value for the ideal gas. The generation energy of the ideal gas of the non-hydrocarbon-based organic compound is obtained through the multiple linear regression-artificial neural network model including the eleventh step of adopting the estimated energy generated by the artificial neural network model as the generated energy value of the ideal gas. How to obtain.
10. The method of claim 9, wherein the optimal molecular descriptors in the third step are independent molecular descriptors of which the values are not the same for all the sample compounds. A method of obtaining the energy of generation of an ideal gas.
10. The method of claim 9, wherein in the fourth step, the training set and the test set are divided by a ratio of 5: 5 to 8: 2. The generation of an ideal gas of the non-hydrocarbon-based organic compound through the multiple linear regression-artificial neural network model How to save energy.
10. The method of claim 9, wherein in the fifth step, the multiple linear regression model searches for the multiple linear regression model by applying a genetic algorithm to the training set. To obtain the generated energy of the ideal gas of the non-hydrocarbon organic compounds.
13. The method of claim 12, wherein the genetic algorithm comprises a population consisting of a plurality of multiple linear regression models randomly drawn from a pool of molecular presenters. Generating; Encoding each individual by combining the numbers of the extracted molecular presenters; Selecting two parent chromosomes from the created population by the Roulette Wheel method and generating offspring by a single point crossover method; A non-hydrocarbon series through a multiple linear regression-artificial neural network model comprising the step of mutating a portion of the chromosomes of the generated offspring with a certain probability and then replacing a portion of the existing population with them to create a new population. A method for obtaining the energy of generation of ideal gas of organic compounds.
The energy generation of the ideal gas of the non-hydrocarbon-based organic compound through the multiple linear regression-artificial neural network model according to claim 9, wherein the fifth step includes determining the predictive performance by the coefficient of determination or the mean absolute error of the regression model. How to obtain.
10. The method of claim 9, wherein the validity in the sixth step is to determine the generated energy of the ideal gas of the non-hydrocarbon-based organic compound through a multiple linear regression-artificial neural network model in which the validity is determined by the t-test value.
10. The method of claim 9, wherein in the eighth step, if the predictive performance of the test set is similar to the predicted performance of the training set, a multiple linear regression model is determined, and the predictive performance of the test set is different from the predicted performance of the training set. After that, the generation energy of the ideal gas of the non-hydrocarbon organic compounds is obtained through the multiple linear regression-artificial neural network model that classifies the training set and the test set again.
10. The method of claim 9, wherein in the ninth step, the search range by the artificial neural network has one hidden layer between the input layer and the output layer and is connected only in a feed forward. A method of obtaining the generated energy of ideal gas of non-hydrocarbon organic compounds through neural network model.
18. The method of claim 17, wherein the activation energy of the hidden layer is obtained by using a sigmoid function to obtain the energy of generation of an ideal gas of the non-hydrocarbon-based organic compound through a multiple linear regression-artificial neural network model. .
10. The method of claim 9, wherein in the tenth step, the overfit prevention reference value is 20. The method for obtaining the generated energy of the ideal gas of the non-hydrocarbon-based organic compound through the multiple linear regression-artificial neural network model.
The method of claim 1, wherein the optimal molecular descriptors extracted in the third step
P ₁ : Energy of Formation for Ideal Gas at 298K from quantum mechanics
P ₂ : Atomic Orbital Molecular Graph Pattern -H1
P ₃ : Atomic Orbital Molecular Graph Pattern-C2
P ₄ : Atomic Orbital Molecular Graph Pattern-C3
P ₅ : Atomic Orbital Molecular Graph Pattern-C4
P ₆ : Atomic Orbital Molecular Graph Pattern-C5
Method for obtaining the energy generated by the ideal gas of the hydrocarbon-based organic compound through a multiple linear regression model comprising a.
The method of claim 9, wherein the optimal molecular descriptors extracted in the third step is
P ₁ : Energy of Formation for Ideal Gas at 298K from quantum mechanics
P ₂ : Atomic Orbital Molecular Graph Pattern-N1
P ₃ : Atomic Orbital Molecular Graph Pattern -N2
P ₄ : Atomic Orbital Molecular Graph Pattern-N3
P ₅ : AOMG Pattern -N4
P ₆ : Atomic Orbital Molecular Graph Pattern -O1
P ₇ : Atomic Orbital Molecular Graph Pattern -O2
P ₈ : Atomic Orbital Molecular Graph Pattern -O3
P ₉ : Atomic Orbital Molecular Graph Pattern-S1
P ₁₀ : Atomic Orbital Molecular Graph Pattern -S2
P ₁₁ : Atomic Orbital Molecular Graph Pattern -S3
P ₁₂ : Atomic Orbital Molecular Graph Pattern-S4
Method for obtaining the generation energy of the ideal gas of the non-hydrocarbon-based organic compound through a multiple linear regression-artificial neural network model.
In the method of obtaining the generated energy of the ideal gas of a hydrocarbon-based organic compound through a multiple linear regression model,
P ₁ : Energy of Formation for Ideal Gas at 298K from quantum mechanics
P ₂ : Atomic Orbital Molecular Graph Pattern -H1
P ₃ : Atomic Orbital Molecular Graph Pattern -C2
P ₄ : Atomic Orbital Molecular Graph Pattern -C3
P ₅ : Atomic Orbital Molecular Graph Pattern -C4
P ₆ : Atomic Orbital Molecular Graph Pattern -C5
Method of obtaining the energy generated by the ideal gas of the hydrocarbon-based organic compound through a multiple linear regression model comprising a.
In the method of obtaining the generated energy of the ideal gas of the non-hydrocarbon organic compound through the multiple linear regression-artificial neural network model,
P ₁ : Energy of Formation for Ideal Gas at 298K from quantum mechanics
P ₂ : Atomic Orbital Molecular Graph Pattern-N1
P ₃ : Atomic Orbital Molecular Graph Pattern -N2
P ₄ : Atomic Orbital Molecular Graph Pattern-N3
P ₅ : Atomic Orbital Molecular Graph Pattern-N4
P ₆ : Atomic Orbital Molecular Graph Pattern -O1
P ₇ : Atomic Orbital Molecular Graph Pattern -O2
P ₈ : Atomic Orbital Molecular Graph Pattern -O3
P ₉ : Atomic Orbital Molecular Graph Pattern-S1
P ₁₀ : Atomic Orbital Molecular Graph Pattern -S2
P ₁₁ : Atomic Orbital Molecular Graph Pattern -S3
P ₁₂ : Atomic Orbital Molecular Graph Pattern-S4
Method for obtaining the generated energy of the ideal gas of the non-hydrocarbon-based organic compound through a multiple linear regression-artificial neural network model comprising a.
A computer-readable storage program for executing the method of obtaining the generated energy of the abnormal gas of the hydrocarbon-based organic compound according to any one of claims 1 to 8, 20, and 22 with a computer. media.
A method for obtaining the generated energy of the abnormal gas of the non-hydrocarbon-based organic compound according to any one of claims 9 to 19, 21 and 23, which can be recorded by a computer program and executed by a computer. Storage media.
A first step of inputting experimental data of radical molecules having an interaction between a resonance structure and a radical center among experimental data of sample organic compounds collected;
A second step of preparing a molecular presenter value for the energy of generation of an ideal gas of the radical molecule having an interaction between the resonance structure and the radical center through the experimental data input in the first step;
A third step of extracting optimal molecular presenters for each experimental data set input in the first step;
A fourth step of separating the experimental data input in the first step into a training set and a test set;
A fifth step of searching for an optimal multiple linear regression model for the training set;
A sixth step of examining the validity of the selected model;
A seventh step of repeating the fifth and sixth steps if there is no validity in the sixth step, and testing the predictive performance of the model with respect to the test set if it is valid;
An eighth step of repeating steps 4 to 7 if the performance does not satisfy the criterion in the seventh step test for the test set, and determining an optimal multilinear regression model if the criterion is satisfied; And
A ninth step of obtaining a generated energy value of an ideal gas through the multilinear regression model searched for the experimental data set of the collected sample organic compounds by the optimal multilinear regression model satisfying the performance test in the eighth step A method of obtaining the generated energy of an ideal gas of a radical molecule having an interaction between a resonance structure and a radical center through a multilinear regression model including a step.
The method of claim 1, wherein in the third step, the optimal molecular descriptor is an independent molecular descriptor whose values are not the same for all the sample compounds. A method for obtaining the energy of generation of an ideal gas of a radical molecule having.
27. The method according to claim 26, wherein in the fourth step, the training set and the test set are divided by a ratio of 5: 5 to 8: 2. A method of obtaining the energy of generation of an ideal gas.
27. The resonance structure of claim 26, wherein the multiple linear regression model searches for the multiple linear regression model by applying a genetic algorithm to the training set in the fifth step. A method for obtaining the energy of generation of an ideal gas of a radical molecule having an interaction between a radical and a radical center.
30. The method of claim 29, wherein the genetic algorithm generates a population consisting of a plurality of multiple linear regression models randomly drawn from the pool of molecular presenters. Making; Encoding each individual by combining the numbers of the extracted molecular presenters; Selecting two parent chromosomes from the created population by the Roulette Wheel method and generating offspring by a single point crossover method; Generating a new population by mutating a portion of the chromosomes of the generated offspring with a certain probability and then replacing a portion of the existing population with them to generate a new population between the resonance structure and the radical center. A method of obtaining the energy of generation of an ideal gas of a radical molecule having interaction.
27. The abnormality of radical molecules having an interaction between a resonance structure and a radical center through a multiple linear regression model according to claim 26, wherein the fifth step includes determining a predictive performance by a coefficient of determination or an average absolute error of the regression model. How to get the generated energy of gas.
27. The method of claim 26, wherein the validity in the sixth step is a multi-linear regression model that determines the validity by the t-test value to determine the generated energy of the ideal gas of the radical molecule having an interaction between the resonance structure and the radical center.
27. The method of claim 26, wherein in the eighth step, if the predictive performance of the test set is similar to the predicted performance of the training set, a multiple linear regression model is determined, and the predictive performance of the test set is different from the predicted performance of the training set. After that, a multilinear regression model that reclassifies a training set and a test set calculates the energy of generation of a standard gas ideal gas of a radical molecule having an interaction between a resonance structure and a radical center.
A first step of inputting experimental data of radical molecules having no interaction between a resonance structure and a radical center among the collected sample organic compounds;
A second step of preparing a molecular presenter value for an energy of generation of an ideal gas of a radical molecule having no interaction between the resonance structure of the sample organic compounds and the radical center;
Extracting optimal molecular descriptors;
A fourth step of separating the experimental data into a training set and a test set;
A fifth step of searching for an optimal multiple linear regression model for the training set;
A sixth step of examining the validity of the selected model;
A seventh step of repeating the fifth and sixth steps if there is no validity in the sixth step, and testing the predictive performance of the model with respect to the test set if it is valid;
An eighth step of repeating steps 4 to 7 if the performance does not satisfy the criterion in the seventh step test for the test set, and separating the three sets after sample normalization if the criterion is satisfied;
A ninth step of searching for an optimal neural network model after dividing the entire sample into three sets;
The absolute value of the difference between the predicted energy produced by the ideal linear regression model satisfying the performance test in the eighth step and the predicted energy produced by the ideal gas calculated by the optimal neural network model found in the ninth step. A tenth step of comparing with a preset overfit prevention reference value; And
If the difference is larger than the reference value for preventing overfitting, the generated energy predicted value of the ideal gas obtained by the multiple linear regression model obtained in the eighth step is adopted as the generated energy value for the ideal gas. Through the multiple linear regression-artificial neural network model including the eleventh step of adopting the generated energy prediction value of the ideal gas by the artificial neural network model as the generated energy value of the ideal gas, there is no interaction between the resonance structure and the radical center. A method of obtaining the energy of generation of an ideal gas of a radical molecule.
35. The method according to claim 34, wherein in the third step, the optimal molecular presenter is an independent molecular presenter whose values are not the same for all the sample compounds. A method of obtaining the energy of generation of an ideal gas of a radical molecule having no interaction.
35. The method of claim 34, wherein the training set and the test set in the fourth step have interaction between the resonance structure and the radical center through the multiple linear regression-artificial neural network model. A method for obtaining the energy of generation of an ideal gas of a radical molecule that does not.
35. The method of claim 34, wherein in the fifth step, the multiple linear regression model searches for the multiple linear regression model by applying a genetic algorithm to the training set. A method for obtaining the energy of generation of an ideal gas of a radical molecule having no interaction between a resonance structure and a radical center.
38. The population of claim 37, wherein the genetic algorithm generates a population of multiple polylinear regression models randomly drawn from a pool of molecular representations. Making; Encoding each individual by combining the numbers of the extracted molecular presenters; Selecting two parent chromosomes from the created population by the Roulette Wheel method and generating offspring by a single point crossover method; Resonance structure and regression through a multiple linear regression-artificial neural network model comprising the step of mutating a portion of the chromosome of the generated progeny (mutation) and then replacing a portion of the existing population with them to create a new population A method for obtaining the energy of generation of an ideal gas of a radical molecule having no interaction between radical centers.
35. The method of claim 34, wherein the fifth step has no interaction between the resonance structure and the radical center through the multiple linear regression-artificial neural network model, which includes determining the predictive performance by the coefficient of determination or the mean absolute error of the regression model. A method for obtaining the energy of generation of an ideal gas of a radical molecule.
35. The energy generated by the ideal gas of the radical molecule having no interaction between the resonance structure and the radical center through a multiple linear regression-artificial neural network model in which the validity in the sixth step is determined by the t-test value. How to obtain.
35. The method of claim 34, wherein in the eighth step, if the predictive performance of the test set is similar to the predicted performance of the training set, a multiple linear regression model is determined, and the predictive performance of the test set is different from the predicted performance of the training set. After that, a multilinear regression-artificial neural network model that reclassifies the training set and the test set is used to obtain the energy of generation of the ideal gas of radical molecules having no interaction between the resonance structure and the radical center.
35. The method of claim 34, wherein in the ninth step, the search range by the artificial neural network has one hidden layer between the input layer and the output layer and is connected only in a feed forward. A neural network model is used to determine the energy generated by an ideal gas of a radical molecule having no interaction between a resonance structure and a radical center.
43. The method of claim 42, wherein a sigmoid function is used as an activation function of the hidden layer, wherein the radical molecules have no interaction between the resonance structure and the radical center through a multiple linear regression-artificial neural network model. A method of obtaining the energy of generation of an ideal gas.
35. The generation of an ideal gas of radical molecules having no interaction between the resonance structure and the radical center through the multiple linear regression-artificial neural network model according to claim 34, wherein in step 10, the reference value for preventing overfitting is 20. How to save energy.
27. The method of claim 26, wherein the optimal molecular descriptors extracted in the third step
P ₁ : Energy of Formation for Ideal Gas at 298K from quantum mechanics
P ₂ : Atomic Orbital Molecular Graph Pattern-C2
P ₃ : Atomic Orbital Molecular Graph Pattern-C3
P ₄ : Atomic Orbital Molecular Graph Pattern-C4
P ₅ : Atomic Orbital Molecular Graph Pattern-C5
P ₆ : Atomic Orbital Molecular Graph Pattern -N2
P ₇ : Atomic Orbital Molecular Graph Pattern -N3
P ₈ : Atomic Orbital Molecular Graph Pattern -O1
P ₉ : Atomic Orbital Molecular Graph Pattern -O2
P ₁₀ : Atomic Orbital Molecular Graph Pattern-S1
P ₁₁ : Atomic Orbital Molecular Graph Pattern -S2
P ₁₂ : Atomic Orbital Molecular Graph Pattern-S3
P ₁₃ : Atomic Orbital Molecular Graph Pattern-S4
P ₁₄ : Balaban Index
P ₁₅ : Area weighted surface charge fraction of hydrogen bond dependent hydrogen bond base atoms (HA dependent HDCA-2 / TMSA)
P ₁₆ : Kier flexibility index
Method for obtaining the energy generated by the ideal gas of the radical molecule having an interaction between the resonance structure and the radical center through a multiple linear regression-artificial neural network model comprising a.
35. The method of claim 34, wherein the optimal molecular descriptors extracted in the third step
P ₁ : Energy of Formation for Ideal Gas at 298K from quantum mechanics
P ₁ : Atomic Orbital Molecular Graph Pattern -H1
P ₂ : Atomic Orbital Molecular Graph Pattern-C2
P ₃ : Atomic Orbital Molecular Graph Pattern-C3
P ₄ : Atomic Orbital Molecular Graph Pattern-C4
P ₅ : Atomic Orbital Molecular Graph Pattern-C5
P ₆ : Atomic Orbital Molecular Graph Pattern-N1
P ₇ : Atomic Orbital Molecular Graph Pattern -N2
P ₈ : Atomic Orbital Molecular Graph Pattern-N3
P ₉ : Atomic Orbital Molecular Graph Pattern -O1
P ₁₀ : Atomic Orbital Molecular Graph Pattern -O2
P ₁₁ : Atomic Orbital Molecular Graph Pattern-S1
P ₁₂ : Atomic Orbital Molecular Graph Pattern-S2
P ₁₃ : Atomic Orbital Molecular Graph Pattern-S3
P ₁₄ : Atomic Orbital Molecular Graph Pattern-S4
Method for obtaining the energy generated by the ideal gas of the radical molecules having no interaction between the resonance structure and the radical center through a multi-linear regression-artificial neural network model.
In the method of obtaining the generated energy of the ideal gas of the radical molecule having interaction between the resonance structure and the radical center through the multiple linear regression model,
P ₁ : Energy of Formation for Ideal Gas at 298K from quantum mechanics
P ₂ : Atomic Orbital Molecular Graph Pattern-C2
P ₃ : Atomic Orbital Molecular Graph Pattern-C3
P ₄ : Atomic Orbital Molecular Graph Pattern-C4
P ₅ : Atomic Orbital Molecular Graph Pattern-C5
P ₆ : Atomic Orbital Molecular Graph Pattern -N2
P ₇ : Atomic Orbital Molecular Graph Pattern -N3
P ₈ : Atomic Orbital Molecular Graph Pattern -O1
P ₉ : Atomic Orbital Molecular Graph Pattern -O2
P ₁₀ : Atomic Orbital Molecular Graph Pattern-S1
P ₁₁ : Atomic Orbital Molecular Graph Pattern -S2
P ₁₂ : Atomic Orbital Molecular Graph Pattern-S3
P ₁₃ : Atomic Orbital Molecular Graph Pattern-S4
P ₁₄ : Balaban Index
P ₁₅ : Area weighted surface charge fraction of hydrogen bond dependent hydrogen bond base atoms (HA dependent HDCA-2 / TMSA)
P ₁₆ : Kier flexibility index
Method of obtaining the generated energy of the ideal gas of the radical molecule having an interaction between the resonance structure and the radical center through a multiple linear regression model comprising a.
In the method of obtaining the generated energy of the ideal gas of the radical molecule having no interaction between the resonance structure and the radical center through the multiple linear regression-artificial neural network model,
P ₁ : Energy of Formation for Ideal Gas at 298K from quantum mechanics
P ₂ : Atomic Orbital Molecular Graph Pattern -H1
P ₃ : Atomic Orbital Molecular Graph Pattern-C2
P ₄ : Atomic Orbital Molecular Graph Pattern-C3
P ₅ : Atomic Orbital Molecular Graph Pattern-C4
P ₆ : Atomic Orbital Molecular Graph Pattern-C5
P ₇ : Atomic Orbital Molecular Graph Pattern-N1
P ₈ : Atomic Orbital Molecular Graph Pattern-N2
P ₉ : Atomic Orbital Molecular Graph Pattern-N3
P ₁₀ : Atomic Orbital Molecular Graph Pattern -O1
P ₁₁ : Atomic Orbital Molecular Graph Pattern -O2
P ₁₂ : Atomic Orbital Molecular Graph Pattern-S1
P ₁₃ : Atomic Orbital Molecular Graph Pattern-S2
P ₁₄ : Atomic Orbital Molecular Graph Pattern-S3
P ₁₅ : Atomic Orbital Molecular Graph Pattern-S4
Method for obtaining the energy generated by the ideal gas of the radical molecules having no interaction between the resonance structure and the radical center through a multiple linear regression-artificial neural network model comprising a.
A computer program for executing a method for obtaining the generated energy of an ideal gas of a radical molecule having an interaction between the resonance structure and the radical center according to any one of claims 26 to 33, 45 and 47. Recordable and computer-readable storage media.
A computer program for executing a method for obtaining the generated energy of an ideal gas of a radical molecule having no interaction between the resonance structure and the radical center according to any one of claims 34 to 44, 46 and 48. Computer-readable storage media.

Images