Skip to content

Day 1

Concepts and Terminologies

What is Machine Learning and why Is It everywhere ?

As the word says Machine and it's Learning means we have a machine or we can say function which takes some input values and predicts or calculates the answer. Easy 🍬🍬 ??

Let's take an example of our brain 🔽

Suppose you are student studying in senior secondary school, and you have a exam in somedays and you need to score good marks 💯 . For these, what you'll do, grab some books and study. By studying, internally your brain is learning pattern that which type of question can occurs, how to solve them. Here comes one term that is training. You are training 🚆🚆 your brain 🧠 to predict or calculate the answers for your upcoming exams. More your brain learns, more it will become accurate to the results. So we discussed is the layman's term for define Machine learning .

To be more technical, A machine 📠 is a system or module which takes data 🗃️ in form of csv, excel file, some documents like docx or pdf, some images or anything. Then we'll clear / filter or preprocess 💇 the data, Then our machine learns crucial patterns 📈 regarding that data like it builds some equations assigning some weights to some parameters. After learning or we can say Model training and after training, it predicts the results for unknown data by applying same pattern that it has learnt before.

Now let's get some familiarity with Machine Learning terminologies -

  • Data 📊 : The raw information or observations used for training and testing machine learning models. Data can be in various forms, such as text, images, numbers, or more.

  • Feature 🔍 : An individual input variable or attribute used in a machine learning model to make predictions or classifications. Features are derived from the data.

  • Label or Target 🎯 : The output or outcome that a machine learning model predicts or classifies. In supervised learning, the model is trained to predict the label.

  • Training Data 📚 : A subset of the data used to train a machine learning model. It includes both the features and their corresponding labels.

  • Testing Data 🧪 : A separate subset of the data used to evaluate the performance of a machine learning model after it has been trained.

  • Algorithm 🧮 : A set of rules and procedures that a machine learning model follows to learn from data and make predictions or decisions. Examples include decision trees, neural networks, and k-nearest neighbors.

  • Model 🤯 : The result of training a machine learning algorithm on data. It represents the learned patterns and relationships in the data.

  • Supervised Learning 👥 : A type of machine learning where the model is trained on labeled data, and its goal is to learn a mapping from input features to output labels.

  • Unsupervised Learning 🤷‍♂️ : A type of machine learning where the model is trained on unlabeled data, and its goal is to discover patterns or structure in the data without specific output labels.

  • Classification 🏷️ : A type of supervised learning task where the goal is to assign data points to predefined categories or classes.

  • Regression 📈 : A type of supervised learning task where the goal is to predict a continuous numeric value as the output.

  • Overfitting 🙅‍♂️ : Occurs when a machine learning model performs well on the training data but poorly on new, unseen data because it has learned noise or irrelevant patterns.

  • Underfitting 🙅‍♂️ : Occurs when a machine learning model is too simple to capture the underlying patterns in the data and performs poorly on both training and testing data.

  • Accuracy ✅ : A common evaluation metric that measures the proportion of correctly predicted instances in a classification task.

  • Loss Function 📉 : A mathematical function that quantifies the error between the model's predictions and the actual target values, used during training to optimize the model.

  • Feature Engineering 🔧 : The process of selecting, transforming, or creating new features from the raw data to improve a machine learning model's performance.

  • Validation Set 🔄 : A subset of the data used during model training to tune hyperparameters and avoid overfitting.

Dive into Supervised Learning

Supervised Machine Learning 🎯 is a type of Machine Learning in which algorithmic model trains on a dataset which have independent features and target both. So in this category of machine learning, model will analyze the patterns and understand how target value can be generated / calculated using independent features.

For example :-

  • Diabetes prediction 🧋 - Suppose we have a dataset having independent features like cholestrol, Blood pressure, systolic pressure etc. and one target feature depicting that person is diabetic or not.

  • House Price Prediction 🏚️ - Suppose we have a dataset having independent features like number of rooms, area (in square km) etc. and one target feature depicting the price of house.

Dive into Unsupervised Learning

Unsupervised Machine Learning 🔴 is a type of Machine Learning in which algorithmic model trains on a unlabelled dataset (means no target) and model will try to identify pattern and create different clusters based on different pattern model observe.

For example :-

  • revenue prediction - An example of unsupervised machine learning would be a case where a supermarket wants to increase its revenue. It decides to implement a machine learning algorithm on its sold products’ data. It was observed that the customers who bought cereals more often tend to buy milk or those who buy eggs tend to buy bacon. Thus, redesigning the store and placing related products side by side can help them understand consumer mindset and increase revenue.

Dive into Reinforcement Learning

Reinforcement Machine Learning 🌪️ is a type of machine learning where an agent learns to make sequential decisions to maximize a cumulative reward. It is commonly used in scenarios where an agent interacts with an environment and learns to take actions that lead to the most desirable outcomes over time.

For example :-

Game playing (e.g., AlphaGo, which learned to play the board game Go), autonomous driving (teaching a self-driving car to navigate safely), robotics (training a robot to perform tasks), and recommendation systems (learning to recommend products or content to users).

Resource 🪵

You can learn more about types of machine learning through this video 📼