stat-215-group

Live colab is here: https://colab.research.google.com/drive/1gmnwWN0POHAz-zj5zxxbLCiFHuW8ZzJr?usp=sharing overleaf: Meeting: Friday 11am to 12pm

Overleaf writing:

• Introduction: Anqi (description of domain problem, climate change, classification problem (clouds)
• EDA: Bodgan
• Feature Selection/Engineering (autoencoder): Finn
• Data Splitting: Bodgan
• Modeling (3 parts): Neural Network & Ensemble (Bodgan),Logistic Regression (Joseph), Random Forest (Anqi)
• Conclusion: Joseph

Bogdan: neural network/Ensemble Joseph: Logistc regression Anqi: Random forest Finn: autoencoder

Data Splits

• train: image1
• validation: image2
• test: image3

Folder Structure:

• code:
  • EDA
  • Modeling

Use the following code to load data and rename the column:

import packages and load the data

• import pandas as pd

• import numpy as np

• import matplotlib.pyplot as plt

• import seaborn as sns

• from time import time

• import os

• from pyreadr import read_r

• image_1 = pd.read_csv("../../data/image_data/image1.txt", delim_whitespace=True, header=None)

• image_2 = pd.read_csv("../../data/image_data/image2.txt", delim_whitespace=True, header=None)

• image_3 = pd.read_csv("../../data/image_data/image3.txt", delim_whitespace=True, header=None)

• column_names = ['y_coor', 'x_coor', 'expert_label', 'NDAI', 'SD', 'CORR', 'Radiance_angle_DF','Radiance_angle_CF','Radiance_angle_BF','Radiance_angle_AF', 'Radiance_angle_AN']

• image_1.columns = column_names

• image_2.columns = column_names

• image_3.columns = column_names