Articles
-
Contest about Tencent Advertisement
I did some attempt to solve the problem. Thought they seems not work well, but I really want to write some methods that may will be used one day.
-
Play CartPole game with Nature-DQN
Recently, people are focusing on the playing of games by computers. There are so many people curious about that why computer score so high and even higher than human. In this blog, we will try to make a smart computer to play CarPole game.
-
Simulation add-operation with pytorch
We all know that the theory Universal approximate theorem, said that a linear layer can simulate all function in theory. But in fact, we can not guaranteed that the training algorithm will be able to learn that function. The first reason is optimization algorithm can not find it. It also may caused by finding wrong function for overfitting.
-
Using tensorflow-lstm predict functions
Lstm nerual network is one kind of recurrent nerual network, and usually used to predict sequences such as language. So we use it to predict some functions such as: sin(x) * k + b, and discuss the factors that influence the accuraccy of lstm.
-
Some Tips for Youngers
I’m 22 years old now, and study in University of Electronic Science and Technology of China. The four years I spended in there are so boring and upseting, so it’s time to rebuild my life and enjoy the further three years.
The reason why I write this blog is mainly for my lover yy, she is so beatiful and lovely but I am so sorry to make her feeling yawnful because of my boring life. I love her, and I will review this blog again and again in the future.
