Laura_Mendomo_Mvomo - Machine Learning Pathway

Concise overview of things learned. Break it up into Technical Area, Tools, Soft Skills
Technical area: I feel more comfortable with word embeddings using BERT
Tools: I gained a more complete understanding of how beautifulsoup works, and learned how to use selenium for bottomless scrolling
Soft Skills: It has been really nice getting to organize my time effectively

Three achievement highlights
Scrapped comments from the Atom forum, played around with my code to improve simplicity/efficiency, did some reading on NLP

List of meetings/ training attended including social team events
I have attended every meeting so far. No social team events.

Goals for the upcoming week
Use the different models with the data collected and check for accuracy.
Refine your strings to be used for the NLP part.

Detailed statement of tasks done. State each task, hurdles faced if any and how you solved the hurdle. You need to clearly mark whether the hurdles were solved with the help of training webinars, some help from project leads or significant help from project leads.
Concatenated all the comments and discussions in one thread as a single data point: no significant hurdle. I used Github (https://github.com/harrycha/Infinite-Scrolling/blob/master/Scroll%20down%20script.py) to get a better understanding of how selenium worked and youtube (https://www.youtube.com/watch?v=ng2o98k983k&t=1137s) for a refresher on beautifulsoup
Read resources on NLP

  • Concise overview of things learned. Break it up into Technical Area, Tools, Soft Skills
    Technical area: Getting more familiar with NLP
    Tools: Learned how to use Doc2Vec
    Soft Skills: Learning how to organize team work when working remotely

  • Three achievement highlights
    Scrapped comments from ketogenic forum, implemented bag of words, learned how to train a model for Doc2Vec

  • List of meetings/ training attended including social team events
    I attended today’s meeting (06/23) and met with my project group

  • Goals for the upcoming week
    Master the fundamentals of BERT. Get the ball rolling on the group project.

  • Detailed statement of tasks done. State each task, hurdles faced if any and how you solved the hurdle. You need to clearly mark whether the hurdles were solved with the help of training webinars, some help from project leads or significant help from project leads.
    Read NLP literature.
    Vectorized the text using Doc2Vec and using my own implementation of bag of words.
    Found similar posts based on Euclidean distance.

Final Self Assessment:

  • Concise overview of things learned. Break it up into Technical Area, Tools, Soft Skills
    Technical area: Worked on classification
    Tools: Learned how to train MLP using sklearn
    Soft Skills: Learning how to organize team work when working remotely
  • Three achievement highlights
    Trained a MLP on classifying each pos by tag, achieved ~40% accuracy on 8 tags (above chance), helped prepare final presentation.
  • List of meetings/ training attended including social team events
    I attended most meetings or gave a heads-up if I had a time conflict.
  • Goals for the upcoming week
    N/A
  • Detailed statement of tasks done. State each task, hurdles faced if any and how you solved the hurdle. You need to clearly mark whether the hurdles were solved with the help of training webinars, some help from project leads or significant help from project leads.
    When choosing classification methods, it was hard to determine which one would be accurate enough while preventing overfitting.