Article Preview
TopIntroduction
Cultivating innovative talents is very important and meaningful for building an innovative society, which has become the consensus of countries all over the world (Hu et al., 2021). Universities are important bases for cultivating innovative talents, and undergraduate education in universities is a key factor in cultivating innovative talent (Zhang, 2020). To improve the quality of undergraduate education, it is necessary to reform and innovate the training mode of undergraduates (Chen et al., 2019).
In recent years, the attempt of improvement of teaching has been continuous, and many innovative educational theories have emerged (Hanandeh et al., 2021). These theories have injected great impetus into teaching reform. Great achievements have been made in the theory and practice of teaching reform, but there is a lack of corresponding research on the quantitative evaluation of reform results. In the whole teaching and learning process, classroom status is an important reference factor for evaluating students' acceptance of courses and teaching quality. In fact, teachers' judgment is the main means of classroom state analysis, but it can distract teachers' attention (Guo, 2021). Therefore, it is important to find a method that can replace teachers in conducting classroom state analysis. In addition, the evaluation of the effectiveness of teaching reform also needs to consider factors such as students' test scores and teaching equipment, and these factors also need to be reasonably evaluated.
In response to the evaluation of the college teaching mode, scholars, and experts have invested a lot of time and energy to carry out some research work and put forward many effective methods of college teaching mode evaluation, which can be roughly divided into: qualitative college teaching mode evaluation method, quantitative college teaching mode evaluation method and the evaluation method of college teaching mode, and the combination of qualitative and quantitative evaluation method of college teaching mode (Harun et al., 2020). However, the evaluation results of the college teaching mode by the qualitative method are highly subjective, and the evaluation results are related to the knowledge background or learning experience (Esson & Wang, 2018).
Different evaluators have different evaluation results in the college teaching mode, and it is easy to add their subjective understanding and assumptions to make the evaluation results not credible (Horne, 2020). Quantitative methods mainly include the evaluation method of the college teaching model of the AHP(Analytic Hierarchy Process), the evaluation method of the college teaching mode of weighted fusion, and the evaluation method of the college teaching model of artificial neural network (McCormack, 2020). Among them, the analytic hierarchy process and weighting method can only describe the fixed change law of the college teaching mode, but the actual college teaching mode has a random change law, so using these methods to evaluate the college teaching mode has a large error (Chen & Lu, 2022). Deep learning is an extension of machine learning research and an effective way to implement artificial neural networks. In recent years, deep learning has become the focus of the field of artificial intelligence. Deep learning theory has achieved excellent results in image recognition, data mining. (Zhang, 2021). There are many implementations of deep learning algorithms, such as multilayer neuron self-encoding neural networks, convolutional neural networks, and deep belief networks. The evaluation results of an artificial neural network in the college teaching mode are better than other methods because it belongs to artificial intelligence technology, has good self-learning ability, and can accurately fit the fixed and random change laws (Zhu et al., 2021). Therefore, the use of deep optimization algorithms to evaluate the effectiveness of teaching reform can overcome the shortcomings of the above methods.