Here’s a cartoon illustration I’ve drawn a while back:
The #machinelearning learning curve
👀🧵👇 See thread below
2/ Starting the learning journey
The hardest part of learning data science is taking that first step to actually start the journey.
3/ Consistency and Accountability
After taking that first step, it may be challenging to maintain the consistency needed to push through with the learning process. And that’s where accountability steps in.
4/ 66 Days of Data
Luckily we have the #66daysofdata by @KenJee_DS that helps us to maintain public accountability. As Ken puts it, you just need to:
-Learn or do #datascience for at least 5 minutes a day
-Tweet or post your progress on social media
5/ Overthinking (Analysis Paralysis)
Sometimes we are overwhelmed with the possibilities and this leads to indecisiveness.
Does this sound familiar?
- R or Python?
- Google or IBM professional certificate in data analytics/data science
6/ Solution to Overthinking
Choose 1 and move on. You can always circle back and give the other a try later on.
7/ Videos I’ve made on:
R vs Python
Google Data Analytics Professional Certificate
8/ Blogs I’ve written on getting started with #datascience
2/ 1. Craft your own personal learning plan
Earlier this year I made a video that details the steps you can take to craft your own personal learning plan for your data journey. Everyone's plan is different, make your own! Here's how...
3/ 2. Work on data projects using datasets that is interesting to you
When starting out, I found that working on datasets that's interesting to you will help you engage in the process. Be persistent and work on the project to completion (end-to-end).
How? Data→Model→ Deployment
Hi friends, here’s my new hand-drawn cartoon illustration ✏️
Quickly deploy #machinelearning models
👀🧵👇 See thread below
2/ Deployment of machine learning models is often overlooked especially in academia
- We spend countless hours compiling the dataset, processing the data, fine tuning the model and perhaps interpreting and making sense of the model
- Many times we stop at that
- Why not deploy?
3/ Topics include:
- Overview of Data science
- Probability and Statistics
- Data cleaning
- Feature engineering
- Modeling
- Classical Machine learning
- Deep learning
- SQL
- Python data structures
2/ Why Do We Need Pandas?
The Pandas library has a large set of features that will allow you to perform tasks from the first intake of raw data, its cleaning and transformation to the final curated form in order to validate hypothesis testing and machine learning model building.
3/ Basics of Pandas - 1. Pandas Objects
Pandas allows us to work with tabular datasets. The basic data structures of Pandas that consists of 3 types: Series, DataFrame and DataFrameIndex. The first 2 are data structures while the latter serves as a point of reference.
1/ #MachineLearning Crash Course by Google
- Free course
- Learn and apply fundamental machine learning concepts
- 30+ exercises
- 25 lessons
- 15 hours to complete
- Real-world case studies
- Explainers of ML algorithms
1/ Interested in how Deep Learning and AI is impacting a 50-year old grand challenge in biology (protein structure folding)?
See this thread 👀🧵👇 #deeplearning#AI#biology#bioinformatics
2/ Deepmind's Alphafold2 Solves Protein Structures (Part 1) #shorts
3/ Deepmind's Alphafold2 Solves Protein Structures (Part 2) #shorts