# DL_practice

**Repository Path**: qyuqxun/dl_practice

## Basic Information

- **Project Name**: DL_practice
- **Description**: FOR ZYB


- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 1
- **Created**: 2022-03-08
- **Last Updated**: 2022-03-11

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Deep learning practice
Congratulations! You have looked through the statistical machine learning including KNN, SVM etc. Now we will come to deep learning. it might be chaos in writing or understanding the network involved, but it still worth a try! Our jounary will begin from CNN and image classfication.

## CNN

CNN might be the mildest deep-learning network you will ever seen. So i will just introduce the dataset we choose and leave the coding to you.

Our dataset is tiny imagenet, a famous set used in CS231N in stanford. The dataset contains 100,000 images of 200 classes (500 for each class) downsized to 64×64 colored images. Each class has 500 training images, 50 validation images, and 50 test images. Wnid.txt contains all the ids and words contains the meaning of these ids. Surely you need to proceed a **convert** to make the directory reasonable.


ENJOY!