Be : - © ficial Intelligence, that use: is and .ights from data, and make* Supervised learning uses labeled data to train the algorithm and make predictions base ) new data. Examples include image classification, speech ition, and spam filtering. Unsupervised learning uses unlabeled data to find patterns and relationships among the data. Examples include clustering and anomaly detection. Reinforcement learning involves training an agent to interact with an environment and learn from rewards and punishments. Examples include game-playing and robotics.* Machine learning requires preprocessing of data, such as cleaning, feature selection malization. * Machine learning models evaluated using various metrics, such as Accuracy, Precision, Recall, and F1-score. * Over Fitting and Under Fitting are common issues in machine learning that require regularization techniquesExample: Image Classification: A machine ng algorithm can be trained to classify images of animals based on their features, such as size, color, and shape. This can be useful in wildlife conservation and identification. Predictive Modeling: A machine learning model can be trained on historical data to predict future trends, such as stock prices, customer behavior, or disease outbreaks. This can help businesses and governments make informed decisions. Natural Language Processing: A machine learning algorithm can be trained to understand and generate human language, such as chat bots, translation, and sentiment analysis. This can improve communication and customer service,a robot to achine learning algorithm can be used to nmendations and advertisements based on es, such as music, movies, and products. This er satisfaction and sales.| TOT Linear regression is a simple and widely used ion problems, where the goal is to predict a based on input features. It finds a linear relationship features and the output variable, and can be used for like the price of a house based on its size._Supervised Learning Algorithm 2. Logistic Regression: parser ee Logistic regression is a binary classification © algorithm that predicts the probability of an input belonging to a certain class, typically usin, 8 @ sigmoid function. It is commonly used in medical diagnosis, spam filtering, and fraud detection. 3. Decision Trees; ———— oa ene Decision trees are a popular algorithm for Classification and regression problems. They work by splitting the dataset into smaller subsets based on the most important features, and recursively building a tree of decisions. They can be used for predicting customer loan approval, and stock prices.Supervised Learning Algorithm a aon Forest: Random forest is a ensemble method that combines: Multiple decision trees to improve the accuracy and robustness of the : _ a redictions. It works by creating multiple decision trees with different subsetS-of the data and features, and aggregating the results. It can be used for predicting customer preferences, credit risk, and disease diagnosis. 8 Faron vector Machines (SVIWI}>——— SVM is a binary classification algorithm that separates the data into two classes by finding the best hyper plane that maximizes the margin between them. It can be used for image classification, sentiment analysis, and fraud detection.Supervised Learning Algorithm 6. Neural Networks : Neural networks are erful and versatile algorithm for solving complex problems, such_as image recognition-natural_language processing, and game playing. They work by simulating the behavior of Reurons in the brain, and can have multiple hidden layers of nodes that ‘extract higher-level features from the input data. Ae, algorithm has its own strengths arf weadpeses, andthe choice depends on the nature of the problem, the size and quality of the data, mieacaasicinrobiem, 4 i ) and the desired level of accuracy and interpretability.Unsupervised Learning Algorithm Unsupervised learning is a type hine learning algorithm where the model learns from unlabelled dais without any specific guidance or supervision. This means that the algorithm is not provided with a predetermined ‘outcome or correct answer to learn from, unlike supervised learning. An example of unsupervised learning is clustering, which involves dividing a dataset into groups or clusters based on similarity. The algorithm looks for patterns in the data that indicate similarities between data points and groups them together. Here are some of the most popular supervised learning algorithms:_Unsupervised Learning Algorithm 2. Custering— Clustering is_a_techs f grouping_similar_data_points together. It can be used to discover patterns in data without any prior knowledge of the groupings. For instance, clustering can be used to segment customers into different ‘groups based on their purchase behavior or demographics. 1. K-Means Clustering Sas oy Hierarchical Clustering Probabilistic Clustering Gaussian Mixture Model Clusteringassociation rule learntag-earbe used to identify the ly purchased items in a supermarket.rning theory provides @sframework to evaluate the learning algorithms and to determine their generalization ent and evaluation of learning algorithms.f selecting theestrelevant-features from the e of the machine learning algorithm. be achieved through Feature Ranking and Subset is the process of ranking the features based on their Za is the procéss of selecting a subset of features that performance of the machine learning algorithm.Statistical Learning Theor methods use SE aac ine learning algorithm to evaluate the ance of the feature Se d methods select the features during the training process of the arnne learning algorithm.Introduction to Statistical Learning Theor SS Evaluating Machine Learning ithms — Evaluating the performance of machine learning algorithms is —ctucial_to determine—their accuracy and generalization Earproperties. Beampaconmonly used pageumance metrics include accuracy, precision, recall, F1-score, and area under the ROC curve. 7 JAXfoss-validation is a technique used to evaluate the Ya performance of machine learning algorithms.Semi Supervised Learning, Reinforcement Learning Markov Detision—Preeess_(IVil Markov Decision Process is a ae ee —mathematical framework—used—to rode! decision-making Processestaay where the outcomes depend on both the current state of the system and the actions taken. MDP is widely used in reinforcement learning to model the interaction between an agent and an environment. me ——_—_§|—__ Bellman Equations - Bellman Equations are a set of recursive equations used to find the optimal value function and optimal policy in a Markov Decision Process. The Bellman equations are used to iteratively update the value function until it converges to the optimal value function.Semi Supervised Learning, Reinforcement Learning —_—__—$ — Markov Decision Process, Bellm: ations, Policy Evaluation, Poli Iteration, and Value Iteration are used to model and solve decision-making problems in Reinforcement Learning. Q-Learning and State-Action-Reward-State-Action (SARSA) are two popular model-free algorithms used to learn the optimal policy in Reinforcement Learning Model-based Reinforcement Learning is a method that uses a model of the environment to learn the optimal policy.2arning is used when the cost of acquiring labeled data feasible to label all the data.ring is an unsupervised learning technique that relies in the form of rewards or penalties.ae) Tolearn fromt | re tomaximizeareward Unsupervised learning with delayed feedback An agent interacts with an environment to learn from trial and error Small amount of labeled data and a large amount of unlabeled data Limited feedback provided by labeled Delayed feedback in the data form of rewards or penalties Image classification with limited labeled _ data, sentiment analysis with limited Game playing, Robotics, _ labeled dataRecommended System Collaborative Filtering — ee ————— * Collaborative Filtering is a_type o/ »mmender system that uses _user~ ec Filtering Her system that uses user” behavior and feedback to make personalized recommendations. — [a The system identifies similarities between users or items and recommends items based on these similarities. 2. Content-basecFiltering — * Content-based Filtering is a type of recommender system that recommends items based on their attributes or features. * The system Qnalyzes the content of the items and recommends similar items to the user. ~n of the human brain. ecognition, image and speech recognition,Recommended System aE Network — . “Multitayer Network is a type of artificia! neural network that has multiple cei =“ * The input layer receives the input data, and the output layer produces the output. * The hidden layers are used to learn complex features and patterns in the data. 6. Back Propagation — Saupaekpropagation is an algorithm used to train artificial neural networks ——— * The algorithm calculates the error between the predicted output and the e actual output and_ adjusts the weights of the network to minimize the error\7. Recommended System Introduction to Deep Learning Deep Learning is a subset of mac -arning that uses artificial neural networks with multiple layers to learn complex features and patterns in the data. Deep Learning is widely used in applications such as image and speech recognition, natural language processing, and autonomous driving,