Beginning the journey with Machine Learning

In the last post, we discussed data science and what all things to learn before solving real-life problems. Today, we’ll be talking about one of the most integral parts of data science, that is, machine learning.

Machine learning (ML) is the scientific study of algorithms and statistical models that computer systems use to perform a specific task without using explicit instructions, relying on patterns and inference instead.

In machine learning, data models are built using machine learning algorithms which take in the data and study patterns forming in the data and give desired results to data scientist, researchers or analysts. The data is divided into two sets, that is, training set and test set. The model is trained over the training set and after training, the model is used to evaluate the test set and give the desired results.

Types of Machine Learning

Supervised Machine Learning

Supervised machine learning is a technique in which models are trained using input-output examples. The input examples consist of the features that are to be analyzed and output consists of the corresponding result derived after analyzing each input example.

There are two types of Supervised machine learning:

• Regression: In this type of supervised learning, the output has continuous value. For example, house prices are a continuous value that depends on the features of the house. The goal of the model is to predict the prices of the houses based on the features provided (no of bedrooms, area, etc.).
• Classification: In this type of supervised learning, the output is in the form of labels (discrete values). In binary classification, the output has two labels, for example, 0/1 or yes/no. In multi-class classification, the output has more than two labels, for eg, a dataset containing four different types of flower pictures and labels are the name of those flowers.

Example of Supervised Learning Algorithms:

• Linear Regression
• Nearest Neighbor
• Gaussian Naive Bayes
• Decision Trees
• Support Vector Machine (SVM)
• Random Forest

Unsupervised Machine Learning

Unsupervised machine learning is a type of machine learning in which inferences are drawn from the input without the help of labels. The output in such cases is not given and the models have to learn on their own, i.e, this type of machine learning supports self-learning.

Unsupervised learning can be further classified into clustering and association

• Clustering: It deals with finding a pattern in a section of uncategorized data. In clustering, the groups are formed of similar data that are called clusters and you can set the number of clusters to be formed by the clustering algorithm.
• Association: Association is where you want to discover rules that describe large portions of your data, such as people that buy X also tend to buy Y.

Some of the unsupervised machine learning algorithms are:

• K-means clustering
• K-NN (k nearest neighbors)
• Dimensionality Reduction
• Principal Component Analysis
• Singular Value Decomposition
• Independent Component Analysis
• Distribution models
• Hierarchical clustering
• Mixture models

Reinforcement Learning

Reinforcement learning is based on the reward system. For every correct prediction, the machine gets a reward and for every wrong answer, the machine gets punished or gets no reward. For example, if a machine correctly identifies the cat picture, it gets 1 else 0. It enables an agent to learn in an interactive environment by trial and error using feedback from its own actions and experiences. In reinforcement learning, the goal is to find a suitable action model that would maximize the total cumulative reward of the agent.

Q-learning and SARSA (State-Action-Reward-State-Action) are two commonly used model-free RL algorithms.

Applications of Machine Learning

Recommender Systems

Shopping websites, social media platforms, and media-service provider companies like Netflix and Amazon Prime are using recommendation systems to provide their customers personalized access to their platform. Shopping websites like Amazon and Flipkart try to understand your likes and dislikes and recommend items according to your liking or it tries you to recommend other user’s choices if you and the other user have the same likes. Netflix and Amazon Prime recommend movies according to your genre of liking or recommend other user’s choices if you and the other user have the same interests.

Virtual Assistant

Google Assistant

Virtual Assistants like Alexa and Google Assistant help in assisting the user to carry out operations like you can ask Google assistant to set a reminder or to set an alarm or to conduct a search. They help us to find and search for things easily and we do not have to waste a large amount of time in surfing. Few of the major applications of Machine Learning here are:

• Speech Recognition
• Speech to Text Conversion
• Natural Language Processing
• Text to Speech Conversion

Self Driving Cars

Google self-driving car

Self-driving cars is one of the most wonderful applications of Machine Learning. Machine Learning has enhanced the driving experience as the driver can sit back while the car automatically takes him to the location. The self-driving car can make turns itself. The safety of the passenger is also enhanced since it keeps monitoring the road for any dangers and takes measures to prevent any accident by reducing the speed of the car or take a turn carefully if there is a deep cut.

Google Translate

Going to new places used to be so difficult since you could not read the boards in different languages. Now, with the help of Google translate, going to new places is not a difficult task anymore.

You can either enter the text and see the conversion into different languages or you can click the picture of the text. Google lens first captures the text in the picture clicked and then uses google translate, which then does text analysis of the given text and translates it into the language of your choice.

Google’s GNMT(Google Neural Machine Translation) is a Neural Machine Learning that works on thousands of languages and dictionaries, uses Natural Language Processing to provide the most accurate translation of any sentence or words. Since the tone of the words also matters, it uses other techniques like POS Tagging, NER (Named Entity Recognition) and Chunking.