Notes from Interviews with Industry Experts
Notes from Interviews with Industry Experts in Data Science, Machine Learning, and Deep Learning.
Photo by Rita Morais on Unsplash
Topic 1: ML & Statistics Foundations
- Inspiration: The very cool blog by Stitchfix
- Statistic & ML Fundamentals:
- work through the book ‘Introduction to Statistical Learning’.
- Elements of Statistical Learning (& its hands-on in R or Python).
- Coursera Course: Andrew Ng’s course on Machine Learning from Stanford
- Software Engineering: work on product deployment: Study on Writeup.ai (e.g., Startup Scripts); an intro is the book ‘Monetize Machine Learning’.
- Udacity Course: Preparing for Coding Interviews
- Public Voice:
- Blogging is useful if you present in a nice way - recruiters may find it useful and intuitive # To Do: Make sure to make videos as they are easy to understand.
- Make scientific contributions.
Topic 2: Coding Language Proficiency
- Become expert in fundamentals by carefully reading ‘Fluent Python’
- Work on Python 101 Challenges (e.g., hacker rank, leet code)
- C: Read the C-Python Article from Stanford
- Additionally: Get proficient in relevant skills: Concurrency, Micro-Services with Django, Redis, Websockets, JS React.
Photo by Chris Ried on Unsplash
Topic 3: NLP - Deep Learning Applications
- Deep Dive into NLP
- Architectures: Understand LSTM and further architectures, read papers on that/Medium articles
- Academic material
- Join MIT classes on NLP / Deep Learning in Winter and Spring Term
- Check for material from Girafski & Chris Manning: look in their research papers / read their books
Photo by Joshua Sortino on Unsplash
Topic 4: Tool Recommendations
- Airflow for Automation/Scheduling (more for the pipeline engineers).
- Domino Data Lab: A platform similar to Github but specialized for data science projects.
- Tableau for exploration with a great competition for people who design dashboards (IronViz).
- Book for ML algorithms is DataSmart because they implement them from scratch, e.g., clustering (using Excel to show what is behind the packages and its math).
- Reminder: MIT squeeze a lot out of it because of all the connections and job opportunities. Focus on doing research.
- Startup: Focus carefully, do not make it take out time too much; may prevent you too much on working on deep learning aspects.
- Kaggle industry perspective: consists of loads of dummy datasets, companies do not necessarily put a lot of weight on Kaggle because data is too clean.
- Focus: Find your focus in the deep learning world like NLP only.
- Potential future role: Forward Deployment Engineer: Consulting & Engineering, e.g. implementing and presenting PoCs.