# Medhini7 - Machine Learning Pathway

**URL:** <http://staging.stemaway.com/t/medhini7-machine-learning-pathway/5582>\
**Category:** Machine Learning Pipeline (NLP)Self-Assessments\
**Created:** [August 5, 2020, 3:31am UTC](http://staging.stemaway.com/t/medhini7-machine-learning-pathway/5582 "2020-08-05T03:31:54Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![medhini7](https://d3rzmptxz7fqzy.cloudfront.net/user_avatar/staging.stemaway.com/medhini7/32/585_2.png) [@medhini7](http://staging.stemaway.com/u/medhini7)\
**Post date:** [August 5, 2020, 3:31am UTC](http://staging.stemaway.com/t/medhini7-machine-learning-pathway/5582/1 "2020-08-05T03:31:54Z")

</div>

**Self Assessment 8/4**

**Technical Area:**

- Web Scraping using BeautifulSoup and Selenium
- Implemented TF-IDF on scraped data
- Understood web embeddings and BERT

**Tools:**

- Jupyter Notebook
- Github
- GSuites
- Slack

**Soft skills:**

- Communicated during meetings to ask leads questions
- Communicated with support group to organize meetings for presentations and to ask questions

**Three achievements:**

- Implemented web scraping on student forum (collected data on 4000+ posts)
- Implemented TF-IDF on the scraped data
- Understood word embeddings and BERT

**List of meetings/trainings attended:**  
7/21 Team Meeting, 7/28 Team Meeting, 7/31 Team Meeting, 8/4 Team Meeting, attended support group meetings as well

**Tasks Completed:**

- Web scraping on student forum
- Applied TF-IDF to the scraped data
- Visualized TF-IDF results using matplotlib

**Goals for upcoming week:**

1. Try to implement word embeddings (I did not get a chance to do this by myself)
2. Implementation of BERT
3. Ask support group more questions when I am confused

**Detailed Statement of Tasks Done:**  
The first week we worked on web scraping, specifically scraping articles from a student forum. We were successfully able to scrape 4000+ posts using code explained during the first team meeting and modifying it. Additionally during the first week, we presented on a high-level overview of BERT to understand the underlying mechanisms that are used. The second week I worked on implementing TF-IDF on the data that was scraped. I did this using the code from office hours as well as other online resources and modifying it to accommodate the large amount of data we were analyzing. From these results, I was able to create a histogram and bar plot of the data using matplotlib. These results were shared during a presentation in today’s team meeting.
