Neda_Sal - Machine Learning Pathway

Concise overview of things learned. Break it up into Technical Area, Tools, Soft Skills:

Technical: I have learned how to scrape a website for data using the Selenium library in Python.

Tools: I have learned how to collaborate using Slack. I am also in the process of learning how to further utilize GitHub.

Soft Skills: I have learned how to effectively communicate technical ideas within my sub-team.

Three achievement highlights:

  1. Successfully scraped a discourse forum, and organized the data efficiently in a pandas DataFrame.

  2. Collaborated with my team and sub-teams to debug all our code, and maximize runtime efficiency.

  3. Cleaned the data, stored it in csv files, and uploaded the csv files along with the corresponding code to my sub-team branch on GitHub.

List of meetings/ training attended including social team event:

All team meetings: 6/1, 6/2, 6/9, 6/13

STEMcasts: Overview of ML and project, Data Mining, Recommendation Models, Git

Training: First Python Training Session

Goals for the upcoming week. Next self-assessment will be due on the following Tuesday 06/23

My goals for the next week are to learn how to use BERT and learn how to use the data we collected to train a model.

Detailed statement of tasks done. State each task, hurdles faced if any, and how you solved the hurdle. You need to mark whether the hurdles were solved with the help of training webinars, some help from project leads or significant help from project leads:

Task 1 Completed: Signed up for Slack, Asana, GSuite. Became familiar with the project goals. Formed sub-teams.

Hurdles: My Gsuite account was not set up. The leads got it set up for me and I was able to login by the next day. The leads split up the whole team into skill-balanced sub-teams.

Task 2 Completed: Wrote code to extract data from the Amazon Seller Discourse forum. The “Account Health” category. We created two DataFrames, one containing the title, category, sub-category, original post content, and the URL. The other contains the URL, and all responses for each post.

Hurdles: I had no prior experience webscraping so I had to teach myself the core concepts. To do this, I re-watched the Data Mining STEMcast and followed along. Jenny, the technical lead, also did a tutorial at one of our meetings where she showed us a scraping example using Selenium. From that I was able to figure out how to use Selenium to scrape data for our project.

Task 3 Completed: Cleaned the data removing any ‘\n’ characters, and any non-English characters.

Task 4 Completed: Stored the clean data in two separate csv files, one for each DataFrame, and uploaded the csv files, and corresponding code to the sub-team’s GitHub branch. Collaborated on code with my teammates through GitHub.

Hurdle: I personally have never used GitHub before, so I was REALLY confused on how to use it. My teammate helped out by creating our branch, and starting by uploading his version of the code. From there, I was able to make necessary changes, and upload an updated version. To get more familiar with GitHub, I re-watched a few clips from the Git STEMcast.

Task 4 Completed: Started keeping a running log to track project progress within my sub-team.

Request change of role if it applies. You may request to become a task lead. Or switch between participant and observer roles.

If my team leads need more support, I would be glad to move up from participant to task lead.

1 Like

Concise overview of things learned. Break it up into Technical Area, Tools, Soft Skills:

Technical: I have learned about BERT conceptually and am in the process of learning how to implement it.

Achievement highlights:

  1. Learned about BERT without having any prior exposure to it.
  2. Took on a role as a task lead.

List of meetings/ training attended including social team event:

All team meetings: 6/16, 6/20

STEMcasts: NLP Webinar 3

Goals for the upcoming week.

Complete the recommendation model.

Detailed statement of tasks done. State each task, hurdles faced if any, and how you solved the hurdle. You need to mark whether the hurdles were solved with the help of training webinars, some help from project leads or significant help from project leads:

Task 1: Learned about BERT

Hurdles: I’ve never had exposure to it before, so I did some googling and also looked at resources my team leads posted.

Task 2: Began organizing the team meetings and project timeline.

Hurdles: I had to learn how to use Asana, and am having trouble fining a time that works for everyone. I am still in the works of finding a good time that works for most participants and leads.

Concise overview of things learned.

I have implemented BERT and am in the process of learning about classification. I am also working on developing my leadership skills.

Achievement highlights:

  1. Finally got my code to work when implementing BERT for the first time.
  2. Gave a solo presentation of the work I have done on the project so far.
  3. Have begun reaching out to the leads start organizing the additional three weeks.

List of meetings/ training attended including social team event:

All team meetings

Goals for the upcoming week.

Create a timeline of project goals for the next three weeks, including deadlines, and schedule of meetings. Also want to make a list of the technologies that the participants will be exposed to/using.

Additionally, I want to catch up on classification of the model and begin looking into improvements to accuracy.

Detailed statement of tasks done. State each task, hurdles faced if any, and how you solved the hurdle. You need to mark whether the hurdles were solved with the help of training webinars, some help from project leads or significant help from project leads:

Task 1: Made a presentation about the ML project and what we’ve accomplished so far.

Task 1 Hurdles: Could not get in contact with my sub-team members regarding the presentation.

Task 2: Implement BERT

Task 2 Hurdles: I struggled a lot with implementing BERT, my code would not run on Jupyter Notebooks, but once I ran it on google colab it worked.

Task 3: Scheduled the final group meetings ahead of time, sent out assignments on Asana and reminders through Slack

Task 3 Hurdles: Had difficulty finding a time that worked for most participants and the project leads. With a bit of planning we overcame most of the time conflicts.