The introductory route of deep learning based on personal experience (simple and fast)
https://blog.csdn.net/weixin_44414948/article/details/109704871
Recently, I have been instructing the introductory of deep learning for my younger brothers and sisters. After they have studied for a period of time, they have arranged the following demo exercises.
1 : By calling sklearn's machine learning algorithm library to implement classification tasks on the fanwei data set , support vector machine (SVM) classification algorithms are required, with an accuracy rate of over 85%. Introduction to fanwei data set: This data set is a txt file of human skeleton nodes of the research group, with a total of 0 and 1 behavior labels.
A better implementation code is as follows: https://blog.csdn.net/weixin_44414948/article/details/109865594
#############################分隔符#######################################
2 : Use the linear regression algorithm to implement the prediction task on the Boston housing price data set . It is required not to call the ready-made linear regression library functions , use numpy to write the linear regression algorithm, and iteratively update the linear regression parameters through the gradient descent method (not directly used Normal equation solving), the accuracy rate is over 85%.
A better implementation code is as follows: https://blog.csdn.net/weixin_44414948/article/details/109867247
#############################分隔符#######################################
3 : Use the deep learning framework (TensorFlow2.0 and above, pytorch1.0 and above) to build a custom convolutional neural network , implement image classification on the locally downloaded imageNet dataset , and require an accuracy rate of more than 70% . (Note: This task mainly exercises the reading of local data sets.)
A better implementation code is as follows: https://blog.csdn.net/weixin_44414948/article/details/109868733