AU2021105525A4 - Mobile user trajectory prediction system with extreme machine learning algorithm - Google Patents
Mobile user trajectory prediction system with extreme machine learning algorithm Download PDFInfo
- Publication number
- AU2021105525A4 AU2021105525A4 AU2021105525A AU2021105525A AU2021105525A4 AU 2021105525 A4 AU2021105525 A4 AU 2021105525A4 AU 2021105525 A AU2021105525 A AU 2021105525A AU 2021105525 A AU2021105525 A AU 2021105525A AU 2021105525 A4 AU2021105525 A4 AU 2021105525A4
- Authority
- AU
- Australia
- Prior art keywords
- trajectory
- mobile user
- machine learning
- sequence
- eml
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
- 238000010801 machine learning Methods 0.000 title claims abstract description 37
- 238000000034 method Methods 0.000 claims abstract description 31
- 238000012549 training Methods 0.000 claims abstract description 19
- VJYFKVYYMZPMAB-UHFFFAOYSA-N ethoprophos Chemical compound CCCSP(=O)(OCC)SCCC VJYFKVYYMZPMAB-UHFFFAOYSA-N 0.000 claims abstract description 5
- 230000000306 recurrent effect Effects 0.000 claims abstract description 5
- 230000006403 short-term memory Effects 0.000 claims abstract description 5
- 230000007774 longterm Effects 0.000 claims abstract description 4
- 238000005070 sampling Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 230000015654 memory Effects 0.000 description 2
- 238000010295 mobile communication Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000003542 behavioural effect Effects 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/0205—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric not using a model or a simulator of the controlled system
- G05B13/026—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric not using a model or a simulator of the controlled system using a predictor
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
- G05B13/027—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using neural networks only
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
- H04W4/029—Location-based management or tracking services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/048—Interaction techniques based on graphical user interfaces [GUI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Abstract
MOBILE USER TRAJECTORY PREDICTION SYSTEM WITH
EXTREME MACHINE LEARNING ALGORITHM
ABSTRACT
The global adoption of smartphones and location-based services has resulted in a
massive and rapid increase in Mobile User data. Because of the large size of Mobile
User data, new possibilities for determining the characteristics of Mobile User mobility
patterns and making mobility predictions emerge. Predicting mobile user mobility is
critical in a variety of modern applications, including personalized recommendation
systems and 5G networks. The present invention disclosed herein is Mobile User
Trajectory Prediction System with Extreme Machine Learning Algorithm comprising of
Trajectory Dataset (101), Extreme Machine Learning (102), Sequence to Sequence
(103), Trajectory Prediction (104); can predict the trajectory of the mobile user with
high accuracy and low mean square error. The present invention disclosed herein uses
Extreme Machine Learning (EML) Algorithm with Sequence to Sequence (Seq2Seq)
Algorithm. The EML with Seq2Seq can predict the trajectories by next locations
prediction with training the trajectories of their previous locations of single or multiple
mobiles users. Predicting location of users plays an important role for 5G Internet
networks as network service providers need to allocate nearest resources to users to
process their mobile request data. The present invention disclosed herein can achieves
good accuracy in predicting the trajectory of the mobile user with low Mean Square
Error (MSE) of 0.00776, compared with the other existing inventions such as Long
Term Short Term Memory (LSTM) in which MSE is 1.85185 and Gate Recurrent Unit
(GRU) with MSE of 11.89521. The present invention, EML with Seq2Seq disclosed
herein is having mobile user prediction accuracy of 95.47%. The Geolife real life
trajectory movement dataset which consist of user's movement latitude, longitude and
users id with each mobile user has 9 locations are considered for training the proposed
present invention.
1/3
MOBILE USER TRAJECTORY PREDICTION SYSTEM WITH
EXTREME MACHINE LEARNING ALGORITHM
DRAWINGS
101 102 103 104
TRAJECTORY EXTREME MACHINE SEQUENCE TO TRAJECTORY
DATASET LEARNING SEQUENCE PREDICTION
Figure 1: Mobile User Trajectory Prediction System with Extreme Machine
Learning Algorithm.
START
UPLOAD DATASET 202
GENERATE EML MODEL 203
r ENTER USER
ID
PREDICT TRAJECTORY
GENERATE MSE GRAPH J.206
STOP
Figure 2: Flow Chart of the present Invention.
Description
1/3
101 102 103 104
Figure 1: Mobile User Trajectory Prediction System with Extreme Machine Learning Algorithm.
UPLOAD DATASET 202
GENERATE EML MODEL 203
r ENTER USER ID
GENERATE MSE GRAPH J.206
Figure 2: Flow Chart of the present Invention.
[0001] The present invention relates to the technical field of Computer Science Engineering.
[0002] Particularly, the present invention is related to Mobile User Trajectory Prediction System with Extreme Machine Learning Algorithm of the broader field of Wireless Communications in Computer Science Engineering.
[0003] More particularly, the present invention is relates to Mobile User Trajectory Prediction System with Extreme Machine Learning Algorithm, in particularly in Wireless Communication Systems. This invention provides a scalable solution for predicting the trajectory of the mobile user with less mean square error and high accuracy by training Extreme Machine Learning with Seq2Seq algorithm.
[0004] Smartphone usage and location-based services have resulted in a significant and quick rise of Mobile User data. Mobile User data presents new opportunities for determining the characteristics of Mobile User movement patterns and developing mobility forecasts. Predicting mobile user mobility is crucial in a variety of modem applications, such as personalized recommendation systems and 5G networks.
[0005] There will be many challenges for mobile networks in the near future, including low network capacity, high latency and limited resources. Maintaining high network performance while also guaranteeing a suitable level of service quality for mobile consumers is critical. To address the issues raised above, intelligent mobility management may be an option. A model of user mobility behaviour will be crucial in this situation for gaining useful information about user behaviour and meeting their needs. Mobile Users generate a large amount of mobility data as a result of their daily activities. These data can be used in a variety of applications to uncover common patterns and gain valuable insights. Many fields, such as mobility-aware services, benefit from this fact. Data mining and learning algorithms can be used to gain a better understanding of users' mobility behavioral patterns and desires in the form of trajectories. As smart phones and location-based services have become more widely adopted around the world, mobility data has exploded in volume and speed. There are new possibilities for identifying mobile user mobility pattern and forecasting its trajectories Fifth-generation (5G) mobile communication systems rely on mobile user trajectory prediction to provide personalized recommendations, intelligent transportation, urban planning, and mobility management. Depending on the application context, the prediction target may differ.
[0006] 5G mobile communications planning relies heavily on forecasting mobile user positions in the near future, which can range from a few seconds to a few minutes in duration. This involves predicting trajectory using a time series of positions separated by a predetermined sampling interval as a reference point. However, when the sampling interval is large, discrete location index trajectory may change between adjacent time steps, whereas when the sampling interval is small, locations may not change between time-steps. This means that the trends in user mobility are not properly reflected. For trajectories based on continuous location coordinates, on the other hand, it is difficult to specify the degree of discretization of the coordinates. Large granularity of discretization allows for the visualization of user movement trends in general. Under high discretization granularity, however, the prediction accuracy may decline as the number of candidate locations grows. To avoid the aforementioned issues, this invention employs a complete method for predicting trajectories made of continuous coordinates with machine learning model. The Extreme Learning Machine (ELM) is a single hidden layer feedforward neural network training technique (SLFN) can predict the mobile user trajectory with high accuracy and low mean square error.
[0007] Referring to Figure 1, illustrates the present invention and main embodiment of current disclosure that is Mobile User Trajectory Prediction System with Extreme Machine Learning Algorithm comprising of Trajectory Dataset (101), Extreme Machine Learning (102), Sequence to Sequence (103), Trajectory Prediction (104); can predict the trajectory of the mobile user with high accuracy and low mean square error.
[0008] The Trajectory Dataset (101) used in the present invention is Geolife real life trajectory movement dataset which consist of user's movement latitude, longitude and users id and each user has nine locations. By training this dataset with Extreme Machine Learning (102) and Sequence to Sequence (103), next sequences of user's locations can be predicted. Earlier algorithms such as KNN and SVM may predict a user's position, but their efficiency degrades as the data size exceeds a certain threshold. To solve this challenge, a combination of advanced EML and Seq2Seq algorithms is employed, which is highly efficient in terms of prediction and processing speed. Each copy of the Extreme Machine Learning (102) has input, output, and forgets cells, and each copy is made up of multiple copies of memory from training data. Input contains input data, output contains output data, and if the outputs are unrelated or the new output is superior to the old output, the old output data is stored in the forgotten cell. This procedure is repeated until all training data has been assigned to input and output cells. In order to train the output cell to predict future location, new test location data will be applied to EML.
[0009] The Extreme Machine Learning (EML) (102) techniques can include Sequence to Sequence (103) algorithms to help predict future location sequences from train data. The Encoder and Decoder are two elements of the Seq2Seq method. The Encoder will turn the training data into a two-dimensional array, which the Decoder will use to predict the future sequences. The Trajectory Prediction (104) mobile user is determined by determining the location of the mobile user automatically in the form of longitude and latitude. The EML with Seq2Seq is trained in such a way that by entering the current user ID, it will predict the future location of the user with longitude and latitude values. The Graphical User Interface (GUI) is designed to facilitate the easy operation and prediction of the mobile user trajectory. The GUI will display the command lanes to be followed and provides the user location in the form of longitude and latitude. The MSE graph is generated to know the MSE values. The MSE values are compared with the existing techniques such as LSTM and GRU. The present invention disclosed herein can achieves good accuracy in predicting the trajectory of the mobile user with low Mean Square Error (MSE) of 0.00776, compared with the other existing inventions such as Long Term Short Term Memory (LSTM) in which MSE is 1.85185 and Gate Recurrent Unit (GRU) with MSE of 11.89521. The present invention, EML with Seq2Seq disclosed herein is having mobile user prediction accuracy of 95.47%.
[0010] The present invention is described in various levels of detail in the Summary of the Invention, as well as the attached sketches and the Detailed Description of the Invention, and the inclusion or omission of components, sections, or other things in this Summary of the Invention is not intended to limit the scope of the present disclosure. For a better understanding of the current disclosure, read the summary of the invention with the thorough description.
[0011] The accompanying illustrations are incorporated into and constitute part of this specification, and they are utilized to better understand the invention. When viewed with the discussion, the drawing depicts exemplary embodiments of the current disclosure and aids comprehension of its concepts. The drawings are solely for illustrative purposes and do not in any way limit the scope of the disclosure. As evidenced by the usage of the same reference numerals, the elements are comparable but not identical. On the other hand, different reference numerals might be used to define linked components. In some embodiments, such elements and/or components may not be present, while in others, they may be present.
[0012] Referring to Figure 1, illustrates the present invention and main embodiment of current disclosure that is Mobile User Trajectory Prediction System with Extreme Machine Learning Algorithm comprising of Trajectory Dataset (101), Extreme Machine Learning (102), Sequence to Sequence (103), Trajectory Prediction (104); can predict the trajectory of the mobile user with high accuracy and low mean square error, in accordance with an exemplary embodiment of the present disclosure to understand the method and the prediction system for mobile user trajectory prediction and accompanied drawing. This illustration is provided to aid comprehension of the disclosure and should not be construed as limiting the disclosure's breadth, scope, or applicability. The invention is not limited to this drawing, and this illustration is provided to aid comprehension of the disclosure and should not be construed as limiting the disclosure's breadth, scope, or applicability. Some elements and/or components, on the other hand, may not be present in embodiments, and others may be used in different ways than those shown in the designs. Depending on the context, the use of a single language to describe a component or element may contain a plural number of such components or elements, and vice versa.
[0013] Referring to Figure 2, illustrates Flow Chart of the present Invention comprising of Start(201), Upload Dataset (202), Generate EML Model (203), Enter User ID (204), Predict Trajectory (205), Generate MSE Graph (206), and Stop (207), in accordance with another exemplary embodiment of the present disclosure and the invention is not limited to this drawing to understand the flow of the present invention to predict the trajectory of the mobile user and MSE graph generation, and this illustration is provided to aid comprehension of the disclosure and should not be construed as limiting the depth, nature, or applicability of the disclosure.
[0014] Referring to Figure 3, illustrates Graphical User Interface (GUI), in accordance with another exemplary embodiment of the present disclosure to easily understand the prediction method and its implementation of the present disclosure, the invention is not limited only to this drawing, and this illustration is provided to assist comprehension of the disclosure and should not be construed as restricting the depth, nature, or applicability of the disclosure.
[0015] Referring to Figure 4, illustrates Black Console with Accuracy Value, in accordance with another exemplary embodiment of the present disclosure to note the Accuracy of the present disclosure, the invention is not limited only to this drawing, and this illustration is provided to assist comprehension of the disclosure and should not be construed as restricting the depth, nature, or applicability of the disclosure.
[0016] Referring to Figure 5, illustrates Mean Square Error (MSE) Graph, in accordance with another exemplary embodiment of the present disclosure to depict the MSE value of the present disclosure, the invention is not limited only to this drawing, and this illustration is provided to assist comprehension of the disclosure and should not be construed as restricting the depth, nature, or applicability of the disclosure.
[0017] The invention will become more well-known as a result of the following detailed discussion, and objects other than those described below will become apparent. The appended drawings are used in this description. The following detailed description of the invention will make the invention more well-known, and objects other than those listed above will become clear. This description refers to the drawings that go with the invention. It is also important to note that additional or alternative measures should be taken. Embodiments are provided so that a skilled person in the art can fully comprehend the current disclosure. In order to provide a thorough understanding of embodiments of the present disclosure, several specifics relating to various components and processes are provided. As those skilled in the art will recognize, the information provided in the embodiments should not be interpreted as limiting the scope of this disclosure. The order of steps revealed in the process and procedure of this invention should not be interpreted as requiring the order defined or illustrated. Additional or alternative steps should be considered as well.
[0018] Referring to Figure 1, illustrates the present invention and main embodiment of current disclosure that is Mobile User Trajectory Prediction System with Extreme Machine Learning Algorithm comprising of Trajectory Dataset (101), Extreme Machine Learning (102), Sequence to Sequence (103), Trajectory Prediction (104); can predict the trajectory of the mobile user with high accuracy and low mean square error. The Trajectory Dataset (101) used in the present invention is Geolife real life trajectory movement dataset which consist of user's movement latitude, longitude and users id and each user has nine locations. By training this dataset with Extreme Machine Learning (102) and Sequence to Sequence (103), next sequences of user's locations can be predicted. Earlier algorithms such as KNN and SVM may predict a user's position, but their efficiency degrades as the data size exceeds a certain threshold. To solve this challenge, a combination of advanced EML and Seq2Seq algorithms is employed, which is highly efficient in terms of prediction and processing speed. Each copy of the Extreme Machine Learning (102) has input, output, and forgets cells, and each copy is made up of multiple copies of memory from training data. Input contains input data, output contains output data, and if the outputs are unrelated or the new output is superior to the old output, the old output data is stored in the forgotten cell. This procedure is repeated until all training data has been assigned to input and output cells. In order to train the output cell to predict future location, new test location data will be applied to EML. The Extreme Machine Learning (EML) (102) techniques can include Sequence to Sequence (103) algorithms to help predict future location sequences from train data. The Encoder and Decoder are two elements of the Seq2Seq method. The Encoder will turn the training data into a two-dimensional array, which the Decoder will use to predict the future sequences. The Trajectory Prediction (104) mobile user is determined by determining the location of the mobile user automatically in the form of longitude and latitude. The EML with Seq2Seq is trained in such a way that by entering the current user ID, it will predict the future location of the user with longitude and latitude values. To preserve the temporal relationship within the trajectory, such as speed or direction, a Seq2Seq framework based on the EML encoder-decoder architecture was constructed. The network uses all paths in a particular area to acquire the mobile user's common short term mobility patterns, which are generated by geographical restrictions.
[0019] Referring to Figure 2, illustrates Flow Chart of the present Invention comprising of Start(201), Upload Dataset (202), Generate EML Model (203), Enter User ID (204), Predict Trajectory (205), Generate MSE Graph (206), and Stop (207), in accordance with another exemplary embodiment of the present disclosure. Initially run.bat file which is the files of invention stored in the local computer need to be click to start (201). After starting (201), the next step is to Upload Dataset (202). The Geolife real life trajectory movement dataset which consist of user's movement latitude, longitude and users id and each user has nine locations is uploaded (202). The Generate EML Model (203) is performed after the Dataset (202) is being uploaded. The EML Model generation is carried by selecting the required number of the layers in the Feedforward Neural Networks. In this model generation, EML is also trained with the Seq2Seq method. The EML model is generated with proper training. The EML Model is generated with given number of features as given by the dataset. The EML machine with Seq2Seq method is initiated or initialized once the model is generated to predict the trajectory of the mobile user. The Enter User ID (204) is used to locate the user location in the form of longitude and latitude. If we enter the user ID as '0.0' now, the EML with Seq2Seq predict the next sequence for user '0' in the form of longitude and latitude. To Predict Trajectory (205) of the mobile user, EML is trained with dataset (202) with longitude and latitude values, the prediction accuracy is found. In predicting the trajectory there will an error in the form of mean square error (MSE). The MSE value can be obtained with generation of MSE graph (206) and then process is stopped (207).
[0020] Referring to Figure 3, illustrates Graphical User Interface (GUI), in accordance with another exemplary embodiment of the present disclosure to easily understand the prediction method and its implementation of the present disclosure. This GUI facilitates the easily understanding of the prediction method and its implementation; an ordinary skilled person also can also train the system with trajectory dataset, can able to predict the trajectory of the mobile user. The GUI front end is designed with the display of the features of the present invention. The modules of the present invention such as Upload Trajectory Dataset, Generate EML Model, Extreme Machine Learning, Predict Trajectory, and MSE Graph are accessible in front end. Once the Uploading of Trajectory Dataset is completed, then Generate EML Model need to be click and it will initiates the training process of the EML with Seq2Seq method, and also other methods such as LSTM (Long Short Term Memory) with Seq2Seq and GRU (Gate Recurrent Unit). After that, when click on the Extreme Machine Learning, command lane will be displayed on GUI related to the EML training process and it can be visualized in black console designed with TensorFlow. This command lane indicates that the training process can be known in black console. By click on the predict trajectory, immediately it will ask for the User Id to predict the next sequence by EML. When we enter the User Id as '0.0' then the next predicted sequence is related to '0'. The predicted sequence of user is indicated with latitude of 7.00911294 and longitude of 5.27950123. The last operation in GUI is MSE Graph generation. By click on this MSE Graph, the MSE (Mean Square Error) of the present invention along with other methods such as LSTM,
GRU are calculated and displayed in GUI.
[0021] Referring to Figure 4, illustrates Black Console with Accuracy Value, in accordance with another exemplary embodiment of the present disclosure to note the Accuracy of the present disclosure. The Black Console is designed with the TensorFlow and it will display the prediction accuracy of the EML with Seq2Seq method. The present invention disclosed here can have the Mobile User Trajectory Prediction Accuracy of 95.47 and Black Console displays the training process.
[0022] Referring to Figure 5, illustrates Mean Square Error (MSE) Graph, in accordance with another exemplary embodiment of the present disclosure to depict the MSE value of the present disclosure. The MSE (Mean Square Error) graph is generated by clicking on MSE Graph on GUI. The present invention disclosed herein can achieves good accuracy in predicting the trajectory of the mobile user with low Mean Square Error (MSE) of 0.00776, compared with the other existing inventions such as Long Term Short Term Memory (LSTM) in which MSE is 1.85185 and Gate Recurrent Unit (GRU) with MSE of 11.89521.
[0023] In order to provide a more detailed understanding of embodiments of the invention; some specific details are laid out in the above exemplary description. An ordinary skilled artisan, on the other hand, might realize that the existing innovation can be implemented without including any of the specific data presented here. For mobile user trajectory prediction, the primary embodiments of the current disclosure are taken into account. The following section describes how the modules are organized on the GUI for easy operation and implementation of the current disclosure. The method and the way of the present embodiment are presented in the above layout for predicting the trajectory of a mobile user with EML trained with Seq2Seq, and it shall not limit the scope of the present disclosure.
Claims (5)
1. Mobile User Trajectory Prediction System with Extreme Machine Learning Algorithm comprising of Trajectory Dataset (101), Extreme Machine Learning (102), Sequence to Sequence (103), Trajectory Prediction (104); can predict the trajectory of the mobile user with high accuracy and low mean square error.
2. Mobile User Trajectory Prediction System with Extreme Machine Learning Algorithm as claimed in claim 1, wherein the Geolife real life trajectory movement dataset which consist of user's movement latitude, longitude and users id with each mobile user has 9 locations are considered for training the EML with Seq2Seq.
3. Mobile User Trajectory Prediction System with Extreme Machine Learning Algorithm as claimed in claim 1, wherein the Extreme Machine Learning (EML) technique is combined with Sequence to Sequence (103) algorithms to help predict future location sequences from train data. The Encoder and Decoder are two elements of the Seq2Seq method. The Encoder will turn the training data into a two dimensional array, which the Decoder will use to predict the future sequences.
4. Mobile User Trajectory Prediction System with Extreme Machine Learning Algorithm as claimed in claim 1, wherein the GUI (Graphic User Interface) front end is designed with the display of the features of the present invention. The modules of the present invention such as Upload Trajectory Dataset, Generate EML Model, Extreme Machine Learning, Predict Trajectory, and MSE Graph are accessible in front end of GUI.
5. Mobile User Trajectory Prediction System with Extreme Machine Learning Algorithm as claimed in claim 1, wherein EML with Seq2Seq can achieves accuracy of 95.47% in predicting the trajectory of the mobile user with low Mean Square Error (MSE) of 0.00776, compared with the other existing inventions such as Long Term Short Term Memory (LSTM) in which MSE is 1.85185 and Gate Recurrent Unit (GRU) with MSE of 11.89521.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| AU2021105525A AU2021105525A4 (en) | 2021-08-15 | 2021-08-15 | Mobile user trajectory prediction system with extreme machine learning algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| AU2021105525A AU2021105525A4 (en) | 2021-08-15 | 2021-08-15 | Mobile user trajectory prediction system with extreme machine learning algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| AU2021105525A4 true AU2021105525A4 (en) | 2021-11-11 |
Family
ID=78479123
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| AU2021105525A Ceased AU2021105525A4 (en) | 2021-08-15 | 2021-08-15 | Mobile user trajectory prediction system with extreme machine learning algorithm |
Country Status (1)
| Country | Link |
|---|---|
| AU (1) | AU2021105525A4 (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114580798A (en) * | 2022-05-09 | 2022-06-03 | 南京安元科技有限公司 | Device point location prediction method and system based on transformer |
| CN114932582A (en) * | 2022-06-16 | 2022-08-23 | 上海交通大学 | Robot small-probability failure prediction method based on Bi-GRU self-encoder |
-
2021
- 2021-08-15 AU AU2021105525A patent/AU2021105525A4/en not_active Ceased
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114580798A (en) * | 2022-05-09 | 2022-06-03 | 南京安元科技有限公司 | Device point location prediction method and system based on transformer |
| CN114580798B (en) * | 2022-05-09 | 2022-09-16 | 南京安元科技有限公司 | Device point location prediction method and system based on transformer |
| CN114932582A (en) * | 2022-06-16 | 2022-08-23 | 上海交通大学 | Robot small-probability failure prediction method based on Bi-GRU self-encoder |
| CN114932582B (en) * | 2022-06-16 | 2024-01-23 | 上海交通大学 | Robot small probability failure prediction method based on Bi-GRU self-encoder |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tekouabou et al. | Improving parking availability prediction in smart cities with IoT and ensemble-based model | |
| He et al. | Graph attention spatial-temporal network with collaborative global-local learning for citywide mobile traffic prediction | |
| AU2021105525A4 (en) | Mobile user trajectory prediction system with extreme machine learning algorithm | |
| Cui et al. | Received signal strength based indoor positioning using a random vector functional link network | |
| Li et al. | Predicting future locations of moving objects with deep fuzzy-LSTM networks | |
| Karatzoglou et al. | A Seq2Seq learning approach for modeling semantic trajectories and predicting the next location | |
| CN110414732B (en) | Travel future trajectory prediction method and device, storage medium and electronic equipment | |
| Kim et al. | Graph convolutional network approach applied to predict hourly bike-sharing demands considering spatial, temporal, and global effects | |
| Li et al. | Pseudo label-driven federated learning-based decentralized indoor localization via mobile crowdsourcing | |
| Xu et al. | Predicting destinations by a deep learning based approach | |
| JP6521835B2 (en) | Movement path prediction device, movement path prediction method, and movement path prediction program | |
| EP3916652A1 (en) | A method and neural network trained by reinforcement learning to determine a constraint optimal route using a masking function | |
| Genders et al. | Policy analysis of adaptive traffic signal control using reinforcement learning | |
| Zhou et al. | Contrastive trajectory learning for tour recommendation | |
| CN114254543A (en) | Prediction using a depth state space model | |
| Zhang et al. | cgail: Conditional generative adversarial imitation learning—an application in taxi drivers’ strategy learning | |
| CN110268338A (en) | It is inputted using vision and carries out Agent navigation | |
| Qin et al. | Next point-of-interest recommendation with auto-correlation enhanced multi-modal transformer network | |
| Xiong et al. | Access patterns mining from massive spatio-temporal data in a smart city | |
| Haddad et al. | Self-growing spatial graph network for context-aware pedestrian trajectory prediction | |
| Shen et al. | Time-aware location sequence recommendation for cold-start mobile users | |
| Lan et al. | Spatio-temporal position-extended and gated-deep network for next POI recommendation | |
| JP2017084079A (en) | Destination prediction apparatus, destination prediction method, and destination prediction program | |
| Qian et al. | Vehicle trajectory modelling with consideration of distant neighbouring dependencies for destination prediction | |
| Ma et al. | Cellular traffic prediction via deep state space models with attention mechanism |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| FGI | Letters patent sealed or granted (innovation patent) | ||
| MK22 | Patent ceased section 143a(d), or expired - non payment of renewal fee or expiry |