Keze Lin

Thinkful Data Science Workshop

1. Name of sponsoring organization: Thinkful
2. Details of the activity (e.g., where, when): Nov 5th  from 8pm -9:30pm. Link: https://app.livestorm.co/thinkful/started-data-science/live?s=d03f66cd-3631-485d-a036-6a86ecedbb34#/chat
3. What you expect to learn:What is data science
   • Collecting all the info in the company and make something valuable
   • Analytics and predictions

   Big Data (Better GPUs)

   • Capture, store, analyze, manage
   • Web 2.0, mobile, wearables, IoT.

   Data science: Venn diagram of Statistics, Computer Science, and Domain Expertise.

   Project pipeline:

   1. Frame the Questions
   2. Collect the data
   3. Process the data
   4. Explore the data
   5. Make predictions
   6. Communicate results

   Data Science use case

   1. Tala, Microfinance
   • The booming middle class in Kenya
   • Little or no access to financial services
   • Task: How to create a score for people who have a little history
   • Asking questions to determine the variables we need.
   • Such as GSM data, spending habits, social networks, price of airtimes, text
   • Use patterns to make a prediction. Classification problem. Decision tree…
   • Communicate to the product, marketing, finance, C-Level.
   1. Industry: Entertainment recommending, Healthcare predicting illness, Agriculture when seed should be planted.

   Data Science Toolkit:

   Foundation: Stat, Linear algebra

   Applications: Python, SQL

   ML:

   Supervised: Classification with discrete outcomes. Regression with a continuous outcome

   Unsupervised: Clustering, grouping data in meaningful ways.

   Environment: Jupyter Notebooks

   We then run a logistic regression model on Jupyter Notebook with python to predict the type of flower by using a pre-loaded Iris data set from Sci-kit learn.

4. How the activity relates to coursework or your career goals: The notes I took above help me to understand data science more.