edlowther / shazam_for_art

Shazam for art

Machine-Learning-Practice

Based on this tutorial

This Practice looks at:

1. Installing the Python and SciPy platform.
2. Loading the dataset.
3. Summarizing the dataset.
4. Visualizing the dataset.
5. Evaluating some algorithms.
6. Making some predictions.

We installed the following:

• scipy
• numpy
• matplotlib
• pandas
• sklearn

1. Import Libraries
2. Load Dataset
3. Get visualisation on the data

Tuesday:

We set out to answer these three key questions:

1. Which feature extraction to do first? HOG, Pixel, Colour Histogram etc
2. What is the order of the scikit pipeline?
3. How does classification work?
   • Do we implement 'supervised'
   • Do we implement 'unsupervised'
4. Imported algorithm from scikit-learn to process an 'unseen' image vs our image bank and return a value. 0 => Lowry, 1 => Turner. Our program successfully assigned the correct value to the unseen image.
5. MVP reached

Wednesday:

We included and compared 3 algorithms to check the artist of a painting. From sklearn we imported DecisionTree, SVM and Neural_Network. The results were as follows:

At a glance svm_pixels appeared to be the most effective algorithm in terms of predicting the correct artist. This was a big break through as we had achieved 50% accuracy.

Thursday:

We tested existing code and created an effective TDD with higher test coverage in our analysis branch. We also researched Django using tutorials and began to build our front end website with Django.

Friday:

We refactored existing code and increased test coverage. We continued with Django progressing into building a simple image upload form

Monday:

We built a Django web app, complete with testing and began implementing some styling We continued to refine the machine learning and improve on our rate of probability

Tuesday:

We set out to complete a few key tasks:

1. Add the remainig features for the app - iRobArt, art movement, art recommendations based on input
2. Styling the front end
3. Connecting the front end with the logic an machine learning
4. Prepare the presentation - to get started and brainstorm ideas