GSOC: Week 10

This weekend I managed to stitch together all three project that have been developed so far.
I.e. I have managed to create a basic example which allows the user to submit a configured logistic regression job to the server, which then fires up a back end Spark job to analyze the data and report back the results

What I did this week:

• Created the Analysis GUI which sends query parameters to the webserver, based on which we build logic of the spark analysis eg. the user would choose some GUI parameters which would be parsed into this kind of query parameters: info_file=/user/digitalAssistanceSystem/data/numbers/infoTrain.txt&fe=dwt-8&train_clf=logreg which would then be parsed and the whole pipeline would be built in Spark
• Added full pipeline functionality to the phase1 Spark analysis application
• Created a REST server which listens for job requests, checks their status and reports back the result
• Tested the whole pipeline as mentioned in the intro

What I aim to do next week:

• Support Decision trees and Random forests classifier configuration
• Test the saving/loading of classifiers
• Create a REST API to get the list of saved classifiers on the server
• Start working on the deep learning (neural netwrk) classifier configuration

Blocks:

• None