CN106875305A - A kind of Teaching quality evaluation method - Google Patents
A kind of Teaching quality evaluation method Download PDFInfo
- Publication number
- CN106875305A CN106875305A CN201611234631.5A CN201611234631A CN106875305A CN 106875305 A CN106875305 A CN 106875305A CN 201611234631 A CN201611234631 A CN 201611234631A CN 106875305 A CN106875305 A CN 106875305A
- Authority
- CN
- China
- Prior art keywords
- model
- teaching
- data
- achievement
- evaluation
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/20—Education
- G06Q50/205—Education administration or guidance
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Educational Administration (AREA)
- Educational Technology (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Electrically Operated Instructional Devices (AREA)
Abstract
The present invention proposes a kind of Teaching quality evaluation method, to solve the deficiency of present colleges Evaluation Method of Teaching Quality, evaluation model is set up using depth learning technology, to the student performance in teaching process, teaching evaluation questionnaires, teaching supervisor's marking table, the conventional performance of student, the conventional quality of instruction level of course to be evaluated, the internal relation of the factors such as the conventional quality of instruction level of teacher is modeled, excavate the central factor for determining quality of instruction, it is institution of higher learning in education activities, personnel training, the decision-making of the aspects such as Staff Building is provided and provided powerful support for.
Description
Technical field
The present invention relates to the research and application, specially a kind of Teaching quality evaluation method of machine learning method.
Background technology
The quality of quality of instruction is an important index for institution of higher learning, and it reflects institution of higher learning's teaching ring
The validity of section, the effective evaluation of quality of instruction is carried out to course, can for institution of higher learning in education activities, personnel training, teacher
The decision-making of the aspects such as money team foundation is provided and provided powerful support for.Institution of higher learning are to the teaching quality evaluation of course by some independent fingers
Mark is constituted, performance of attending class, total marks of the examination, the evaluation to course teaching, evaluation of teaching supervisor of such as student etc., but these
Index has subjective index, also there is objective indicator, can mix some human factors in wherein, is only obtained from these index weighted arrays
Index can not accurately reflect real quality of instruction；And some quantitative targets are brought into clear and definite quality of instruction calculating
In formula, often to a great extent left and right Faculty and Students make great efforts direction, the methods of learning aid, to teaching work
Significant guide effect is played, can deviate the basic goal of education activities.Current researcher uses machine learning
With the method for statistical learning from the initial data of teaching process, the quality of instruction to course is evaluated.Document " Hu Shuai,
The such as Gu Yan, Qu Weiwei are based on teaching quality evaluation scale-model investigation [J] the Henan science of PCA-LVQ neutral nets, 2015, (7):
1247-1252. " proposes a kind of teaching being combined based on principal component analysis (PCA) and learning vector quantization neural network (LVQ)
Environmental Evaluation Model.Teaching quality appraisal system is set up using analytic hierarchy process (AHP) (AHP), then is initially commented with principal component analysis extraction
The characteristic information of valency index system, will be input to LVQ neutral nets by the characteristic information after dimension-reduction treatment.The method is level
Analytic approach is combined with vector quantization neutral net, is extracted the rule in teaching quality appraisal system, and use neutral net
It is modeled.The problem of the method is to need to carry out principal component analysis to each index of existing index system, and can not be complete
Accomplish to directly obtain the fine or not degree of quality of instruction by initial data entirely, and influenceed by neural metwork training cost is big,
It is difficult to extract the central factor in teaching quality evaluation, and the situation of overfitting compares in the case of sample size is less
Seriously, a model that can reflect universal law is hardly resulted in.
The weak point of existing method is：
(1) most methods are, based on fixed formula, to be imparted knowledge to students by being weighted to the index in teaching process
The quantitative assessment of quality, many times this can not reflect the level of real quality of instruction, while also for education activities bring
Negative guidance quality effect；
(2) existing methods seldom consider student for the validity that teaching process feeds back, and course history to be evaluated
On the factor such as achievement level, it is difficult to quality of instruction is carried out effectively and really to evaluate.
Based on this, the present invention proposes a kind of Teaching quality evaluation method, to solve present colleges teaching quality evaluation
The deficiency of method, evaluation model is set up using deep learning technology, to the student performance in teaching process, teaching evaluation questionnaires,
The conventional teaching of the conventional quality of instruction level of the conventional performance of teaching supervisor's marking table, student, course to be evaluated, teacher
The internal relation of the factors such as quality level is modeled, and excavates the central factor for determining quality of instruction, is that institution of higher learning are being taught
The decision-making of the aspects such as activity, personnel training, Staff Building is provided and provides powerful support for
The content of the invention
In order to overcome the weak point of existing Teaching quality evaluation method, including existing method is based on fixation mostly
Formula, the quantitative assessment of quality of instruction is obtained by being weighted to the index in teaching process, existing method seldom considers
To student for the validity that teaching process feeds back, and the factor such as the historical achievement level of course to be evaluated, the present invention is specially
Profit proposes a kind of Teaching quality evaluation method based on deep learning.Feature of the invention includes teaching evaluation data
Pretreatment, the design of deep learning model, model training, the calculating of teaching quality evaluation result, each process include several
Step, its feature is described as follows respectively：
(1) pretreatment of teaching evaluation data
The teaching evaluation data that the present invention is directed to include three kinds, respectively student achievement data, Students' evaluation data, teaching
Inspection evaluating data, the pretreatment mode of the data of three types is as follows：
C. student achievement data
The record of every student of each subject one, the data processed in the present invention for " final examination achievement " and " usually into
The two achievements are standardized by achievement ", the value for making it be converted into [0,1] interval.Standardized method is：
- for the achievement of fraction type, such as 0-100/ achievement, it is using method directly divided by full marks, such as full
It is 100 points to divide, and the fraction of certain student is 90 points, then the fraction of the student is 90/100=0.9 points after standardizing；
- for grade type achievement, such as A, B, C, D, E five, it is 1 point by best grade, minimum grade is 0 point, its
Remaining each grade 0 and 1/be uniformly distributed；
D. Students' evaluation data
Students' evaluation data are questionnaire type, are filled in after student completes the study of certain subject, including from teaching
The aspects such as content, teaching material, teacher are evaluated.In general for each problem or a basket of questionnaire, have
Corresponding weights, last religion score is commented for calculating.In the present invention using the initial data of teaching evaluation questionnaires, do not consider
Weights.Questionnaire is made up of a series of multiple-choice question, and the processing mode to questionnaire data is dummy variable in the present invention, for example,
For certain multiple-choice question, there are 4 options, selection of certain student to the topic is 1 and 2, then dummy variable turns to 1100.
The exercise question that the word in questionnaire is answered is not considered.
E. teaching supervisor's evaluating data
It is evaluation to classroom instruction situation that teaching supervisor evaluates, including to the content of courses, Teaching Skill, situation of preparing lessons,
Interactive, student listen to the teacher state etc. evaluation, each single item evaluation content typically in the form of grade type, such as it is excellent, good, in and
Lattice, difference, or use hundred-mark system form.
- for the evaluation of fraction type, such as 0-100/ achievement, it is using method directly divided by full marks, such as full
It is 100 points to divide, and the fraction of a certain evaluation is 90 points, then the fraction that this evaluates after standardizing is 90/100=0.9 points；
- evaluated for grade type, it is such as excellent, good, in, pass, differ from five, be 1 point by best grade, minimum grade is
0 point, remaining each grade 0 and 1/be uniformly distributed；
(2) design of deep learnings model
A. the design of achievement-comment religion model
Student's a branch of instruction in school fraction that achievement-comment religion model receives to pre-process by step (1) (including final examination into
Achievement and usual performance) and religion data are commented as input to the course, it is output as student's up to the present all course achievements
Ranking grade, one is divided into p grade, and such as one has 100 students, 10 grades, then entitled 1st grades of 1-10,11-20
It is the second grade, the rest may be inferred.Specifically, the input layer of the model is a vector, and the first half of vector is student
A branch of instruction in school is warp of the student to the course by " the final examination achievement " and " usual performance " after standardization, latter half
That crosses after standard comments religion vector, and both connect into an input vector；The output of the model is that a length is that p and its value exist
Real-valued vectors between [0,1], after an output is often obtained, it is 1 to set maximum of which component, and remaining position is 0, a certain position
For 1 representative model exports grade of the student representated by this, in each output vector, have and only one is 1, remaining
Position is 0.If there is largest component arranged side by side in certain output, random selection one is 1, and remaining is disposed as 0.
Model is made up of, wherein in addition to last block, each block is wrapped using the structure of deep neural network multiple blocks
Full articulamentum, active coating and a screen layer are included, last block is made up of full articulamentum and active coating.
- full articulamentum
Unification in the present invention is the full articulamentum for being originally inputted 3 times of layer vector length using node number, such as defeated
Incoming vector is 50 dimensions, then the nodal point number of each full articulamentum is 150, each node in full articulamentum and last layer
All nodes have a connection, and each connection has the real number weights between [0,1], in each node of full articulamentum
Place performs the operation of weighted sum；
- active coating
In each full articulamentum followed by an active coating, the node number of active coating and the node number of full articulamentum
Identical, the node of the node upper full articulamentum of correspondence of each active coating is carried out to the output valve by full articulamentum node
The calculating of sigmoid functions, i.e.,：Wherein x is upper one output of full articulamentum correspondence node, and y is active coating phase
Answer the output of node；
- screen layer
In each active coating followed by a screen layer, screen layer shields a node of the g% of active coating at random,
The node for being shielded is no longer participate in the calculating of next full articulamentum.Just by randomly selected side before model starts training
Formula determines that each screen layer needs the node for being shielded.
B. the design of teaching quality evaluation model
The input of teaching quality evaluation model includes two parts, one is evaluating number by the teaching supervisor after standardization
According to, the second is the last blocks of active coating output characteristic of achievement-the comment religion model to each teaching class per subject is counted,
It is a q dimensional vector, what it was obtained by：The final examination achievement of this class each student and the student to the course
Comment religion information input achievement-comment religion model after, obtain the last blocks of active coating output characteristic vector of the model,
By each student obtain this vector each component be added, then again each component divided by the teaching class student people
Number.
Teaching quality evaluation model is output as the numerical value between a 0-1, represents the scoring of quality of instruction.
Model is made up of using the structure of deep neural network full articulamentum, active coating and screen layer, these three types
Layer is identical with the A points of this part, and model structure is identical with the A points of this part.
(4) model trainings
The weights of model are adjusted using the error back propagation learning algorithm of deep neural network standard, adjustment is
What the difference between output and real desired value according to model was carried out.
For the error calculation of achievement-comment religion model, the 0-1 vectors according to output and the student by historical data calculating
The difference of grade vector (computational methods are shown in step (2) .A) is carried out, if two vectors are identical on a certain position, the error of this
It is 0, otherwise is 1.
For the error calculation of teaching quality evaluation model, the quality of instruction scoring exported by model and by teaching management people
Difference between the quality of instruction scoring that member is given is carried out.
The weights of model are initialized using the random number between [0,1], carry out many wheel training, and all training samples are defeated
Enter in model and complete weighed value adjusting for a wheel, untill the output error of model no longer declines.
(5) teaching quality evaluations
The step of certain subject to certain teaching class carries out teaching quality evaluation is as follows：
D. obtained according to step (1) and pre-process student achievement data, Students' evaluation data, teaching supervisor's evaluating data；
E. to every student of the class that imparts knowledge to students, calculated according to step (2) .B the q dimensional features that are obtained from achievement-comment religion model to
Amount；
F. the q dimensional feature vectors for obtaining are attached with the teaching supervisor's evaluating data after standardization, are input to teaching
Environmental Evaluation Model, model is output as the fraction between [0,1], and the fraction reflects teaching class on the subject
Quality of instruction.
Specific embodiment
The present invention is tested during the teaching quality evaluation of colleges and universities where inventor, achieves preferable effect.
One embodiment is given below, teaching quality evaluation is carried out to " operating system " course in the network engineering specialty of the colleges and universities.
(1) data prediction
Teaching class is the student of Third school grade last term, and one has 98 students, complete in the learning process of the first two years
Into 18 subjects.The final examination achievement and usual performance of course " operating system " to be evaluated are 100 points of systems, by invention
The step of appearance (1) .A is standardized.
Students' evaluation data are that every subject fills in a questionnaire per student, and questionnaire includes 21 road single choice test items and 1 road
Word answer is inscribed, and single choice test items are 4 options, the step of by the content of the invention (1) .B to every student for 21 road individual events
The answer of multiple-choice question carries out dummy variable, and the one 84 0-1 vector of dimension is obtained after dummy variable.
Teaching supervisor's evaluating data is that every subject fills in a questionnaire, and questionnaire includes 15 evaluations of aspect, wherein having 8
Individual aspect need to be given 0-100/ evaluation, in addition 7 aspects need to be evaluated at five grades in the middle of selected, by invention
The step of content, (1) .C was standardized to this partial data, and one 15 real-valued vectors of dimension, each component are obtained after standardization
Value between [0,1].
(2) design of deep learning model
A. the design of achievement-comment religion model
The step of by the content of the invention (2) .A is carried out, and wherein p takes 20, g and takes 20.The input dimension of model is tieed up for 2+84=86.
The output dimension of network is 20 dimensions.Table 1 illustrates the specific design scheme of achievement-comment religion model.
The achievement of table 1.-comment religion modelling table
B. the design of teaching quality evaluation model
The step of by the content of the invention (2) .B is carried out, and defeated dimension is tieed up for 258+15=273, wherein the data of 258 dimensions come from into
The output characteristic (being shown in Table the 36th layer of 1) of last active coating of achievement-comment religion model.The output dimension of network is 1 dimension.Table 2
Illustrate the specific design scheme of teaching quality evaluation model.
The teaching quality evaluation modelling table of table 2.
(3) model trainings
MatConvNet (http://www.vlfeat.org/matconvnet/) in table 1 is realized by configuration file
With the model structure in table 2, data set is made Matlab data file .mat forms, then provided using MatConvNet
Training script cnn_train.m is trained.Training carries out 70 wheels, and learning rate is respectively 0.05,0.01,0.005,0.001,
0.0005,0.0001 with 0.00005.The loss function of training uses zero-one loss.Each model is trained by 70 wheels
Afterwards, system can generate 70 .mat files, and the parameter of model at the end of each wheel is trained is saved respectively.
(4) teaching quality evaluations
Achievement-the comment religion model at the end of the 70th wheel training is first used to the total marks of the examination and Students' evaluation of " operating system "
The vector of data composition carries out feature extraction, extracts the defeated of last active coating after the data input to model of each student
Go out (being one 258 vector of dimension), each component of vector as 98 is averaged, after then this is passed through averagely
258 dimensional vectors be attached with by teaching supervisor's evaluating data of " operating system " after standardization, be input into the 70th training in rotation
Teaching quality evaluation model at the end of white silk, finally gives the fraction between a 0-1, exists multiplied by teaching class with 100, is represented
Quality of instruction on the course.
Claims (5)
1. a kind of Teaching quality evaluation method based on deep learning, it is characterised in that the described method comprises the following steps,
(1) pretreatment of teaching evaluation data；
(2) design of deep learning model；
(3) model training；
(4) calculating of teaching quality evaluation result.
2. the method for claim 1, it is characterised in that the pre-treatment step of the teaching evaluation data is specifically included：
Teaching evaluation data are obtained, the teaching evaluation data include student achievement data, Students' evaluation data and teaching supervisor
Evaluating data, the pretreatment mode for the data of the three types is as follows：
A. student achievement data
The record of every student of each subject one, the data processed in the present invention are " final examination achievement " and " usual performance ",
The two achievements are standardized, the value for making it be converted into [0,1] interval；Standardized method is：
- use the direct method divided by full marks for the achievement of fraction type；
- for grade type achievement, be 1 point by best grade, minimum grade is 0 point, remaining each grade 0 and 1/
Equally distributed method treatment；
B. Students' evaluation data
Students' evaluation data are questionnaire type, and the processing mode to questionnaire data is dummy variable；
C. teaching supervisor's evaluating data
Teaching supervisor's evaluation is the evaluation to classroom instruction situation, and hundred-mark system form is used in the form of grade type；
- for the evaluation of fraction type, using method directly divided by full marks；
- evaluated for grade type, be 1 point by best grade, minimum grade is 0 point, remaining each grade 0 and 1/
It is uniformly distributed.
3. method as claimed in claim 2, it is characterised in that the design procedure of the deep learning model is specifically included：
A. the design of achievement-comment religion model
Achievement-comment religion model to receive student's student achievement data and Students' evaluation data by step (1) pretreatment as defeated
Enter the input vector of layer；The output of the model is that a length is the real-valued vectors of p and its value between [0,1], is obtained often
After being exported to one, it is 1 to set maximum of which component, and remaining position is 0, and a certain position exports the student to be somebody's turn to do for 1 representative model
Grade representated by position, in each output vector, has and only one is 1, and remaining position is 0；If being deposited in certain output
In largest component arranged side by side, then it is 1 to randomly choose one, and remaining is disposed as 0；
The model is made up of, wherein in addition to last block, each block is wrapped using the structure of deep neural network multiple blocks
Full articulamentum, active coating and a screen layer are included, last block is made up of full articulamentum and active coating：
- full articulamentum
It is the full articulamentum for being originally inputted 3 times of layer vector length using node number, is performed at each node of full articulamentum
The operation of weighted sum；
- active coating
In each full articulamentum followed by an active coating, the node number of active coating and the node number phase of full articulamentum
Together, the node of the node upper full articulamentum of correspondence of each active coating, is carried out to the output valve by full articulamentum node
The calculating of sigmoid functions, i.e.,：Wherein x is upper one output of full articulamentum correspondence node, and y is active coating phase
Answer the output of node；
- screen layer
In each active coating followed by a screen layer, screen layer shields a node of the g% of active coating at random, is shielded
The node for covering is no longer participate in the calculating of next full articulamentum, just true by randomly selected mode before model starts training
Fixed each screen layer needs the node for being shielded,
B. the design of teaching quality evaluation model
The input of teaching quality evaluation model includes two parts, one is by the teaching supervisor's evaluating data after standardization, its
Two is that the last blocks of active coating output characteristic of achievement-the comment religion model per subject to each teaching class is counted, and it is
One q dimensional vector, what it was obtained by：The final examination achievement of this class each student and the student are commented the course
After religion information input achievement-comment religion model, the last blocks of active coating output characteristic vector of the model is obtained, by every
Individual student obtain this vector each component be added, then again each component divided by the teaching class number of student；
Teaching quality evaluation model is output as the numerical value between a 0-1, represents the scoring of quality of instruction；
Teaching quality evaluation model is made up of using the structure of deep neural network full articulamentum, active coating and screen layer, and this three
It is identical in the layer of type and achievement-comment religion model.
4. method as claimed in claim 3, it is characterised in that the model training step is specifically included：
Using the error back propagation learning algorithm of deep neural network standard to teaching quality evaluation model and achievement-comment religion mould
The weights of type are adjusted, and adjustment is carried out according to the difference between the output of model and real desired value：
For the error calculation of achievement-comment religion model, the 0-1 vectors according to output and the student's grade by historical data calculating
The difference of vector is carried out, if two vectors are identical on a certain position, the error of this is 0, otherwise is 1；
For the error calculation of teaching quality evaluation model, the quality of instruction scoring exported by model with by teaching manager to
Difference between the quality of instruction scoring for going out is carried out；
The weights of model are initialized using the random number between [0,1], carry out many wheel training, and all training samples are input to
In model and weighed value adjusting is completed for a wheel, untill the output error of model no longer declines.
5. method as claimed in claim 4, it is characterised in that the teaching quality evaluation step is specifically included：
A. obtained according to step (1) and pre-process student achievement data, Students' evaluation data, teaching supervisor's evaluating data；
B. to every student of the class that imparts knowledge to students, the q dimensional feature vectors obtained by achievement-comment religion model are calculated according to step (2)；
C. the q dimensional feature vectors for obtaining are attached with the teaching supervisor's evaluating data after standardization, are input to quality of instruction
Evaluation model, model is output as the fraction between [0,1], and the fraction reflects teaching of the teaching class on the subject
Quality.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611234631.5A CN106875305A (en) | 2016-12-28 | 2016-12-28 | A kind of Teaching quality evaluation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611234631.5A CN106875305A (en) | 2016-12-28 | 2016-12-28 | A kind of Teaching quality evaluation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106875305A true CN106875305A (en) | 2017-06-20 |
Family
ID=59164097
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611234631.5A Pending CN106875305A (en) | 2016-12-28 | 2016-12-28 | A kind of Teaching quality evaluation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106875305A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784883A (en) * | 2017-12-01 | 2018-03-09 | 重庆电子工程职业学院 | Tutoring system and the Evaluation Method of Teaching Quality based on the tutoring system |
| CN108596420A (en) * | 2018-02-02 | 2018-09-28 | 武汉文都创新教育研究院（有限合伙） | A kind of talent assessment system and method for Behavior-based control |
| CN108734370A (en) * | 2017-12-27 | 2018-11-02 | 上海储翔信息科技有限公司 | A kind of intelligent curriculum points-scoring system excavated based on machine learning, big data |
| CN108764718A (en) * | 2018-05-28 | 2018-11-06 | 王春宁 | Selection method, system are estimated and volunteered to college entrance examination score based on deep learning algorithm |
| CN110334810A (en) * | 2019-07-10 | 2019-10-15 | 福州大学 | MOOC corpse course recognition methods based on machine learning |
| CN111915227A (en) * | 2020-08-21 | 2020-11-10 | 广州云蝶科技有限公司 | Teaching evaluation data analysis method |
| CN111915225A (en) * | 2020-08-21 | 2020-11-10 | 广州云蝶科技有限公司 | Teaching evaluation data processing method |
| CN112667776A (en) * | 2020-12-29 | 2021-04-16 | 重庆科技学院 | Intelligent teaching evaluation and analysis method |
| CN115689820A (en) * | 2022-09-27 | 2023-02-03 | 东南大学附属中大医院 | Learning quality evaluation method based on two-way and continuous medical education closed-loop management system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105931116A (en) * | 2016-04-20 | 2016-09-07 | 帮帮智信（北京）教育投资有限公司 | Automated credit scoring system and method based on depth learning mechanism |
| CN105976098A (en) * | 2016-04-28 | 2016-09-28 | 大连理工大学 | BP neural network-based college and university teaching quality evaluation method |
| CN106227851A (en) * | 2016-07-29 | 2016-12-14 | 汤平 | Based on the image search method searched for by depth of seam division that degree of depth convolutional neural networks is end-to-end |
-
2016
- 2016-12-28 CN CN201611234631.5A patent/CN106875305A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105931116A (en) * | 2016-04-20 | 2016-09-07 | 帮帮智信（北京）教育投资有限公司 | Automated credit scoring system and method based on depth learning mechanism |
| CN105976098A (en) * | 2016-04-28 | 2016-09-28 | 大连理工大学 | BP neural network-based college and university teaching quality evaluation method |
| CN106227851A (en) * | 2016-07-29 | 2016-12-14 | 汤平 | Based on the image search method searched for by depth of seam division that degree of depth convolutional neural networks is end-to-end |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784883A (en) * | 2017-12-01 | 2018-03-09 | 重庆电子工程职业学院 | Tutoring system and the Evaluation Method of Teaching Quality based on the tutoring system |
| CN108734370A (en) * | 2017-12-27 | 2018-11-02 | 上海储翔信息科技有限公司 | A kind of intelligent curriculum points-scoring system excavated based on machine learning, big data |
| CN108596420A (en) * | 2018-02-02 | 2018-09-28 | 武汉文都创新教育研究院（有限合伙） | A kind of talent assessment system and method for Behavior-based control |
| CN108764718A (en) * | 2018-05-28 | 2018-11-06 | 王春宁 | Selection method, system are estimated and volunteered to college entrance examination score based on deep learning algorithm |
| CN110334810A (en) * | 2019-07-10 | 2019-10-15 | 福州大学 | MOOC corpse course recognition methods based on machine learning |
| CN111915225A (en) * | 2020-08-21 | 2020-11-10 | 广州云蝶科技有限公司 | Teaching evaluation data processing method |
| CN111915227A (en) * | 2020-08-21 | 2020-11-10 | 广州云蝶科技有限公司 | Teaching evaluation data analysis method |
| CN111915227B (en) * | 2020-08-21 | 2024-03-19 | 广州云蝶科技有限公司 | Teaching evaluation data analysis method |
| CN111915225B (en) * | 2020-08-21 | 2024-03-19 | 广州云蝶科技有限公司 | Teaching evaluation data processing method |
| CN112667776A (en) * | 2020-12-29 | 2021-04-16 | 重庆科技学院 | Intelligent teaching evaluation and analysis method |
| CN112667776B (en) * | 2020-12-29 | 2022-05-10 | 重庆科技学院 | Intelligent teaching evaluation and analysis method |
| CN115689820A (en) * | 2022-09-27 | 2023-02-03 | 东南大学附属中大医院 | Learning quality evaluation method based on two-way and continuous medical education closed-loop management system |
| CN115689820B (en) * | 2022-09-27 | 2023-08-15 | 东南大学附属中大医院 | Bidirectional learning quality assessment method and continuous medical education closed-loop management system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106875305A (en) | A kind of Teaching quality evaluation method | |
| Papić et al. | Identification of sport talents using a web-oriented expert system with a fuzzy module | |
| Golle et al. | Effectiveness of a “grass roots” statewide enrichment program for gifted elementary school children | |
| CN108256102A (en) | A kind of Independent College Studentss based on cluster comment religion data analysing method | |
| Cwik et al. | Not feeling recognized as a physics person by instructors and teaching assistants is correlated with female students’ lower grades | |
| Satria et al. | Application of SAW in the Class Leader Selection Decision Support System | |
| Saltiel | What’s math got to do with it? multidimensional ability and the gender gap in stem | |
| Lawn | Voyages of measurement in education in the twentieth century: Experts, tools and centres | |
| Clark et al. | Field trips in natural resources professional education: The Panama case and recommendations | |
| Adak et al. | An elective course suggestion system developed in computer engineering department using fuzzy logic | |
| Retrialisca et al. | Decision Support System and Recommendation on SBMPTN Try-Out with Analytic Hierarchy Process (AHP) | |
| WO2022217113A1 (en) | Systems and methods for learner growth tracking and assessments | |
| Symeonaki et al. | Markov systems with fuzzy states for describing students’ educational progress in Greek Universities | |
| Liu et al. | Prediction and assessment of student learning outcomes in calculus a decision support of integrating data mining and Bayesian belief networks | |
| Conner | Acknowledging the elephant in the room: A multiple-case study exploring the experiences of social studies teacher-coaches | |
| Jeganathan et al. | Predicting Academic Performance of Immigrant Students Using XGBoost Regressor | |
| Han | The construction of the structure model of college students' positive mental qualities | |
| Halil | Testing Multistage Testing Configurations: Post-Hoc vs. Hybrid Simulations | |
| do Valle Silva Andrade et al. | Modelling with self-organising maps and data envelopment analysis: a case study in educational evaluation | |
| Cwik et al. | How the learning environment predicts male and female students' motivational beliefs in algebra-based introductory physics courses | |
| Aru et al. | Assessment Of The Admission Criteria That Predict Students' Academic Performance in Undergraduate Years in a Nigerian University | |
| Yang | Internship effect prediction for physical education majors based on artificial neural network | |
| Wang et al. | Design of Learning Environment for Undergraduate Comprehensive Literacy Education under Blended Learning Environment | |
| KAYNAK et al. | Student Consultancy Service: Prediction of Course Grades in Course Selection Phases Using Artificial Intelligence Techniques | |
| Guo | Construction and Analysis of Computer Model in Sports Education Effect Evaluation Based on Grey System Theory |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170620 |