How to Get Started?

Hey there! If you’re starting to explore this blog, you came to the right place.

You could get started with the machine learning basics if you need to, and then dive into deep learning for image recognition, text processing, and more!

If you’re into data science in general, you will also find below articles about python, databases, visualization, and accelerated computing.

Python for machine learning

• Python Crash Course for Machine Learning : The absolute minimum needed about python to get started with machine learning.
• Numpy Crash Course for Machine Learning : Efficient data array manipulation with python.
• Matplotlib for Machine Learning : Basic visualization with a small machine learning example.
• Fighting Bugs in Python : We all write bugs. Here is some advice on how to squash them.

Machine learning basics

• Handwritten Digit Recognition with scikit-learn : simple introduction to a dense neural network, and to the scikit-learn library
• First Neural Network with Keras : same exercise, this time with keras
• The 1-Neuron Network: Logistic Regression : how is a neuron working?
• Logistic Regression vs Neural Network: Non Linearities : what is a non-linear problem, and how to deal with it.
• Overfitting Illustrated : overfitting might be the most important issue in machine learning. See it with your own eyes

Deep learning

Image recognition

• Tuning a Deep Convolutional Network for Image Recognition with Keras and TensorFlow (Handwritten Digits) : learn how to adapt the parameters of a convolutional network for best performance
• Image Recognition: Dogs vs Cats (92%) : discriminate pictures of dogs and cats, with special emphasis on data augmentation and the choice of the optimizer.
• Image Recognition with Transfer Learning (98.5%) : use state-of the art pre-trained models for easy and powerful image classification. Introduction to the VGG and ResNet neural networks.
• Object detection with darknet : use the darknet program and pre-trained networks to find and identify objects in many categories in a picture
• Real-time human detection with OpenCV : detect humans in a video with the HOG algorithm. That’s not really deep learning but still useful!

Natural Language Processing

Learn how to train a machine to understand the meaning of text.

• Simple Text Mining with Pandas (yelp dataset part 1) : pandas is immensely useful for exploratory analysis. Try it in basic text analysis.
• Text Preprocessing For Machine Learning (yelp dataset part 2) : how to turn text into arrays of numbers for a neural network
• Word Embedding and Sentiment Analysis (IMDB) : turn words into multidimensional vectors, and actually understand what the heck this means
• Sentiment Analysis with a deep convolutional network (yelp dataset part 3) : convolutional networks are not only for image processing. Use such a network to predict if a text is positive or negative

Unsupervised learning

• Visualizing Datasets: look at your data before you do anything with it. Learn about dimensionality reduction with t-SNE.

Data science

• COVID-19 Analysis : early analysis of the COVID-19 data with python, holoviews, matplotlib.
• COVID-19 Analysis: Uncertainties: A proper treatment of uncertainties for COVID-19

Databases

• MongoDB in 10′ : Fill with python, Read with pandas : MongoDB is a cool NoSQL database that can make your life easier. Set up your server, fill it with python, and process the data with pandas
• MongoDB : Remote Access (raspberry pi) : remote access to a MongoDB server, with a simple introduction to firewall management.

Visualization

• Matplotlib for Machine Learning : How to do your first plots
• Interactive Visualization with Bokeh in a Jupyter Notebook : Matplotlib is not for interactive display. Try bokeh!
• Data pipeline with Docker, InfluxDB, and Grafana : full data pipeline for IOT, experimental science, or the industry. You’ll get an introduction to excellent tools: docker for virtualization, InfluxDB as a time-series database, and Grafana as a multi-purpose dashboard server.
• Show your Data in a Google Map with Python : study real-estate price in France near Geneva on a Google Map, with pandas and bokeh.
• Choropleth Maps in Python : Learn how to do choropleth maps (colored maps with political or physical boundaries) avec geoviews et geopandas.

IOT

• Fast Minimal Install of a Raspberry Pi : Standard installation of raspbian Lite, to get only what you need.
• Raspberry pi Zero W : Headless Install : Headless installation of Raspbian Lite, without screen and without keyboard.

Accelerated computing

• Numpy Crash Course for Machine Learning
• Make python fast with numba : compile your python code easily for your CPU
• Boost python with your GPU (numba+CUDA) : compile your python code for your nvidia GPU
• CUDA kernels in python : Dive into details, and write custom CUDA kernels for your GPU in python

Tools

• Install TensorFlow on Windows for Deep Learning
• Install TensorFlow on Linux for Deep Learning
• Install Anaconda for Machine Learning and Data Science in Python
• Remote jupyter notebooks with ssh port forwarding
• Backup with rsync on Windows (WSL)

Scientific blogging and web

• Welcome to the New Blog! : A short description of what I did to convert the blog to Django.
• Jupyter Notebooks in Web Pages : Finally manage a perfect integration of jupyter notebooks on your website.