Machine Learning using Python
What is Machine Learning
Machine learning (ML) is a subset of Artificial Intelligence (AI) that often uses statistical techniques to give computers the ability to "learn" with data, without being explicitly programmed.
Machine learning is nothing but to recognise patterns in your data.
Machine Learning is everywhere!
From automatic recommendations of movies to watch, to what food to order, which products to buy to personalized music/video playlist and recognizing the people in the photos, most of the websites, mobile phones and other devices around us have machine learning at their core. Complex websites and applications from every tech giant contains multiple machine learning models.
Google alone has majority of its servies backed by advanced ML system and few are listed below
- Google Search Engine
- Google Translate
- Google Maps
- Google Ads
- YouTube
- Google Voice Recognition System
- Google Assistant
- Google Drive
- Gmail
Other tech giants that are known for their ML technology include
- Amazon
- Netflix
- Baidu
- IBM
- Microsoft
- Intel
- Apple
- Uber
- Spotify
Machine Learning can solve our problems
Beyond commerical applications, ML has a tremendous influence on the way data-driven researsh is done now. From finding distant planets, discovering scientific particles, forecasting weather, recognizing handwritting, detecting creditcard faradulent, identifying spam emails, identifying equipment anamoly, detecting DDoS attacks, preventing security intrusions, providing personalised medical treatment and diagnosis, Machine learning and Deep Learning can be used to solve wide range of problems.
Why Machine Learning?
It's hard to handwrite programs to develop the intelligence required to solve the problems mentioned earlier. Think of your life without an intelligent spam filter. We could still hardcode a list of words to look in each mail and blacklist them as "Spam". However, they have 2 main disadvantages:
- The logic required to make a decision is very specific to single domain and task. Changing the task even slightly might require rewritting the entire system.
- Designing those logic rules requies an expert level understanding on the problem we are trying to solve.
These kind of carefully handcoded expert level decision systems will definitely fail in detecting faces from an image in our phone. Infact, it's an extremely difficult task even for experts to identify the best characteristis needed to detect human face.
Using ML, simply presenting the program with a large collection of images of faces is enough for an algorithm to determine what characterstics are needed to identify a face.
To find more on Why machine learning is unavoidable?
, read this link.
ML Prerequisities
Familiarity with one of the programming languages such as Python, R, MatLab
& basic knowledge of Linear Algebra, Matrix, Vectors, Probability & Statistics.
Over here, we will use Python as our core programming language.
Why Python? Packages, Packages everywhere !
- Want to work in text -
nltk, numpy, scipy, scikit-learn, tensorflow - Want to work in audio -
librosa - Want to solve machine learning problem -
pandas, scipy, scikit-learn, tensorflow - Want to see the data clearly -
matplotlib, seaborn, scikit - Want to use deep learning -
tensorflow, pytorch - Want to do scientific computing -
scipy - Want to integrate web applications -
Django - Want to talk to hardware -
Pyserial - Want to create GUI/ Apps -
PyQT5, Kivy, Tkinter - Want to work with DB -
SQLite
Why Python is great choice in my 👀?
- Clear, readable syntax
- General Purpose
- Interpreted -> Rapid Prototyping
- Quick progression
- Versatility
- Widely available resources
- Supportive community
Topics covered
- Anaconda
- Jupter Notebook
- Python Refresher
- Numpy introduction
- Numpy introduction
- Our First ML Model
- K Nearest Neighbors
- Linear Regression
- Logistic Regression