Self-Assessment for Machine learning Module 1
- Get familiar with the machine learning workflow and other programs as apart of of module.
- Go over the whole process for building the recommender system, and more details for project expectations.
- Grasp a basic understanding of the software used for this internship.
- Practice communication and interaction with group members.
- Practice using the stemaway interface.
- Setup all coding environments as a part of this project.
- Think deeper into the Machine learning workflow and grasping a better understanding of the project.
- Gained a better understanding of beautiful soup and selenium as a direct result of the first module’s content.
Self-Assessment for Module 2:
Things that I learned
- Utilized web scraping tools to gather data from different websites.
- Accesed different files (mainly csv) and how to scraped data from them
- Compared scraped data obtained from sites
- Utilized jupyter notebooks & am becoming more comfortable using it
- Developed familiarity with scraping data from the sites
- Successfully scraped data from one of the forums
- Successfully scraped data from a simpler website after Successfully doing so from a forum
- Cleaned and pre-procssed data
- Performed analysis of data obtained and of course scraped the data.
Self-Assessments Module 3:
Web scraping of forum data
Python usage increased
I learned teamwork by asking for help when I began to scrape the forum data. I used the help I got to proceed with scraping and exporting data.
Loaded scraped data and exported it in an excel file.
Explored scraped data to gain insights.
Visualized the data in excel tabs.
I got stuck while trying to scrape my category from the car forum. Eventually, after trial and error, I was able to figure out what to do and proceeded to explore the data.
Self Assessment for Module 4:
- Learned how to train BERT
- Was able to combine various models like BERT with other models to assess their overall effectiveness and their accuracy.
- Learned how BERT operates and how it can function
- Simple Transformers
- Jupiter Notebook *Pytorch
- Was able to train various models and see what their relative efficiencies were.
- Tried using different types of data, such as cleaned and not cleaned, to see which one was more effective in producing results (and again overall accuracy).
- Tried to dock the ML App with models
- As the modules became progressively more complex, it became a bit harder to understand the “meat” of the modules, so having a better understanding of machine learning in its entirety is a plus. I had a lot of trouble at first, but I was eventually able to obtain some of the knowledge I needed to complete this module’s tasks…