Workshop by Navya Prakash

10/12/19 - Link to repo

Data Science (Extract and making human sense)

• Domain Expertise
• Statistics
• Coding

Roles

• Data Analys
• Data Engineer
• Data Scienctist
• Data Architect
• Developer

11/12/19 - Link to repo

80% spent cleaning and prep data

if age, median > mean then more on the higher side

What is Machine Learning? -automates data extrapolation and classification

Prediction

• linear regression

Classification

• logistic regression
• decision tree

Predictive & Preemptive Analytics

• self sustain algo

As a person, we have gathering data all the time

Predictive ML model

1. clean
2. filter
3. build
4. train
5. test

Type of ML

• supervised learning
• unsupervised learning (groupings, recommendation)

could platform

• Quick and affordable
• repeated use

Microsoft Azure ML algo cheat sheet

Capstone project!!

julia(like python but faster), R

12/12/19 - Link to repo

split different routes for GET and POST so no duplicate request if refresh

2 components

• the tranlation (AI to translate)
• which part of the image (AI to retrieve the text)

Azure doc

api

• breaks if more than 10 people
• can adjust threshold

check out

• azure cognitive services apis
• how-old.net
• what-dog.net (broken)
• microsoft learn labs

AI ethics

• AI data and privacy considerations (public info are plaintext, critical data must be encrypted)

how data is stored, belong to group (what is the size of groups, legal restrictions varies)