search

datalass1 / fastai

this repo will show code and notes covered during the fastai course
0 stars 0 forks source link

lesson-3 homework to work on kaggle competitions for Imagery data and Structured data #17

Closed datalass1 closed 5 years ago

datalass1 commented 5 years ago

Using Multilabel classification create a notebook for Understanding the Amazon from Space competition.

Using Structured data approach with some Machine learning techniques create a notebook for the Rossmann Store Sales competition.

datalass1 commented 5 years ago

Planet Understand the Amazon from Space

Kaggle competition

Step 1: Download the data

$ pip install –U kaggle-cli
$ kg config –u <your username> -p <your password> -c <competition name>
$ kg download

At this point I had to increase my HD from 50GB to 100GB Check out HD space using: $ df -h total Paperspace requires that the partition is updated with the new storage expand the disk after storage upgrade: $ sudo /var/lib/cloud/scripts/per-instance/ps_resize.sh

Extract data: zip files $ unzip –q <filename.zip> Extract data: tar files $ 7za x <filename.tar.7z> This extracts 7z format and delivers an output tar xf <filename.tar>

Step 2: Set up a Jupyter notebook and follow lesson 3 1hr20min-2hr section.

See notebook, lesson3-planet_understanding_the_amazon_from_space-20190120.ipynb, in fastai repo

Step 3: Optional: Consider a haze removal technique

CVPR2009 Best Paper award Single Image Haze Removal using Dark Channel Prior - Kaiming He, Jian Sun, and Xiaoou Tang

datalass1 commented 5 years ago

HINT: in the fastai library do not edit existing notebooks, otherwise unable to pull to update the repo. Create copies to run/edit and making my own notebooks, data, folders should not impact a git pull.

datalass1 commented 5 years ago

Rossmann Competition

Step 1. Get the Data

Jeremy provides a link in his lesson3-rossmann notebook, include wget at the beginning, the data is accessible: $ wget http://files.fast.ai/part2/lesson14/rossmann.tgz Unzip the files: $ tar -xvzf rossmann.tgz x for extract v for verbose z for gnuzip f for file, should come at last just before file name -C will allow contents to unzip to a different directory tar -xvzf /path/to/yourfile.tgz -C /path/where/to/extract/ source

datalass1 commented 5 years ago

completed amazon from space, did not complete the Rossmann competition because......A NEW FASTAI came out :) v3 2019 so I have made new issues to work through the newest version of the MOOC.