https://github.com/gaussalgo/mlp_2018

Machine Learning Prague 2018 workshop repository.
https://github.com/gaussalgo/mlp_2018

Last synced: 13 days ago
JSON representation

Machine Learning Prague 2018 workshop repository.

• Host: GitHub
• URL: https://github.com/gaussalgo/mlp_2018
• Owner: gaussalgo
• Created: 2018-03-21T13:35:26.000Z (almost 7 years ago)
• Default Branch: master
• Last Pushed: 2019-12-23T14:24:43.000Z (about 5 years ago)
• Last Synced: 2024-11-08T12:34:35.251Z (2 months ago)
• Language: Python
• Size: 15.6 KB
• Stars: 0
• Watchers: 6
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# # MLP_2018

## Task 00 - Initialization

* File `00_init.py`

* Install requirements (may need to restart session)

* Download data

## Task 01 - Image classification

* open `01_image_classification.py`

* go through the process description comments

* run the script

* check that it works 

## Task 02 - Image similarity search

* open `02_image_similarity.py`

* go through the process description comments

* run `02_image_similarity.py` on single thread (as is)

* check results

## Task 03 - Parallelization of image preprocessing

* open `03_01_is_master.py`

* go through the process description comments

* run `03_01_is_master.py` on 3 threads (workers)

* run `02_image_similarity.py`

* check results and compare speed of processing with task 02

## Task 04 - Flask server and similarity of uploaded image upload and similarity

* create teams

* add a simple flask web app to the project that can upload new images

* export the server to CDSW_PUBLIC_PORT (port 8080 - https://www.cloudera.com/documentation/data-science-workbench/latest/topics/cdsw_web_ui.html)

* run the web server and open the link from the top right hand corner dropdown

* copy URL and send to your smart phone

* upload a picture of your shoe :)

* update path in `02_image_similarity.py`

* check which shoes are similar