standing-o/Data_Analysis_Club_Assignment

Programming Assignment - Kaggle Competition, Python quiz, etc.

Data Analysis Club Assignment

• Activities in data analysis club - Programming Assignment
• May 14, 2020 ~ Dec 1, 2021

Description

0. Python Quiz for newcomer | Quizs
1. Supervised Learning : MNIST Handwriting Recognition
2. Unsupervised Learning : IRIS data Clustering
3. Kaggle Tutorial : Data Analysis and Machine Learning with POKEMON stat & image dataset
4. Kaggle Competition : Predict Future Sales
5. Kaggle Competition : Plant Seedlings Classification

Assignment 1. Supervised Learning : MNIST Handwriting Recognition

• May. 29, 2020

Presentation | Code1, Code2

Assignment 2. Unsupervised Learning : IRIS data Clustering

• This project is maintained by 최연석, 오서영
• Jul. 23, 2020

| Presentation | Code

Assignment 3. Kaggle Tutorial : DA and ML with POKEMON stat & image dataset

• Aug. 14, 2020

| Presentation | Code

Dataset

[1] Pokemon with stats, https://www.kaggle.com/abcsds/pokemon  
[2] Pokemon Image Dataset, https://www.kaggle.com/vishalsubbiah/pokemon-images-and-types  

Pokemon generation with GANs

• Vanilla GAN with Keras
• Convert RGBA to RGB (PNG to JPG)

Assignment 4. Kaggle Competition : Predict Future Sales

• This project is maintained by 최연석, 오서영
• Aug. 28, 2020

| Presentation | Code1, Code2

Competition

[1] Predict Future Sales, https://www.kaggle.com/c/competitive-data-science-predict-future-sales

Results

1. LSTM

RMSE error : 1.02245 (10 iterations)
RMSE error : 1.02711 (60 iterations)

2. XGBRegressor

RMSE error : 1.43613

Assignment 5. Kaggle Competition : Plant Seedlings Classification

• This project is maintained by 오서영, 허지혜
• Sep. 18, 2020

| Presentation | Code1, Code2

Competition

[1] Plant Seedlings Classification, https://www.kaggle.com/c/plant-seedlings-classification/overview

Results

1. Baseline CNN

Train accuracy: 72.50%
Val accuracy: 69.38%

2. Resnet

Train accuracy: 86.25%
Val accuracy: 80%