ব্যাচের কার্যক্রম - মেইন লাইভ ক্লাস শুরুর আগেই যারা ইনরোল করেছেন, তারা কি বসে থাকবেন ? না! এই মডিউলে ব্যাসিক বেশ কিছু ভিডিও দেওয়া আছে। শুরু হয়ে যাক তবে শেখার জার্নি! 

Objective: Understand the Full ML Pipeline and Get ready to deep dive into the world of ML

Live Class 1: Master Pandas, Numpy , Vector and its Importance in ML

Live Class 2: Machine Learning Pipeline , Best Practices | Underfit , Overfit , Bais , Variance

Objective: Understand the full course and get a breif knowledge about ML

Live Class 1: AI Engineer Full Roadmap | Intro to ML

Live Class 2: Types of Machine Learning | Intro to EDA

Objective: Introduce Git, Jupyter, Colab, and environment setup.

Live Class 1: What is Git and why use it | Basic Git commands: init, add, commit | Understanding version control | What is GitHub and how it works | Creating and managing repositories on GitHub | Cloning and pushing code | Branching and merging basics | Typical project folder structure | Best practices for organizing codebase

Live Class 2: What is Jupyter Notebook | Running and managing code cells | Introduction to Google Colab | Uploading and sharing notebooks | Writing clean and readable code | Using comments and proper indentation | Writing Markdown in notebooks | Formatting text, lists, and code blocks with Markdown | Creating headings and notes for better understanding

Live Class 1: Introduction to linear regression | Fitting a line to data using least squares | Understanding slope and intercept | What is overfitting in regression | Introduction to regularization | Ridge regression and L2 penalty | Lasso regression and L1 penalty | Comparing ridge and lasso | When to use ridge vs lasso

Live Class 2: Introduction to model evaluation | What is Mean Absolute Error (MAE) | Interpreting MAE in regression models | What is Root Mean Squared Error (RMSE) | Difference between MAE and RMSE | What is R-squared (R²) | Interpreting R² as model accuracy | Choosing the right metric for evaluation

Project: House Price Predictor

Live Class 1: Introduction to K-Means clustering | How K-Means groups similar data points | Choosing number of clusters | What is the Elbow Method | Using Elbow Method to find optimal K | Introduction to PCA (Principal Component Analysis) | Reducing dimensionality with PCA | Interpreting principal components | PCA vs clustering: when to use what

Live Class 2: Hierarchical Clustering & DBSCAN

Project: Clustering Assignment: Iris Dataset

Live Class 1: Introduction to Convolutional Neural Networks (CNN) | What is computer vision? Use cases in industry | Basics of CNNs: convolution, pooling, flatten, FC | CNN architecture overview: input, convolution, pooling, fully connected layers | What is a convolution layer and its purpose | Applying filters to input images | Understanding kernel size and stride | Max pooling and average pooling layers | Reducing spatial dimensions with pooling | Role of convolution and pooling in feature extraction

Live Class 2: Introduction to image classification | Overview of CNN for image classification | Using Keras for building CNN models | Loading and preprocessing image data | Defining the CNN architecture in Keras | Adding convolution, activation, and pooling layers | Flattening and adding fully connected layers | Compiling the model with optimizer, loss, and metrics | Training the model with image data | Evaluating model performance on test data | Image preprocessing techniques: resizing, normalization | Data augmentation for Image Data

Project: Dog vs Cat Image Classifier

Live Class 1: Building Scalable ML APIs with FastAPI
What is FastAPI and why it's ideal for ML deployment | Setting up a FastAPI app for serving predictions | Creating robust endpoints with Pydantic for validation | Serving ML models using joblib or pickle | Structuring APIs: /predict, /health, /info endpoints | Testing APIs using Swagger UI, ReDoc, and Postman | Intro to async handling and real-time performance advantages

Live Class 2: Dockerizing Your ML Service & Intro to MLflow
Docker Focus: | What is Docker and why use it for ML APIs | Writing a Dockerfile for FastAPI apps | Building Docker images and running containers | Exposing containerized APIs over localhost/port | Testing containerized predictions with curl/Postman | Analyzing API responses and troubleshooting | Basics of cloud deployment | Deploying Dockerized applications on cloud platforms | Using services like AWS, Heroku, or Google Cloud for deployment

Project: Production Ready ML Model Pipeline with FastAPI, Docker & MLflow

Live Class 1: Introduction to logistic regression | Understanding sigmoid function and probabilities | Binary classification with logistic regression | Introduction to K-Nearest Neighbors (KNN) | How KNN makes predictions based on distance | Choosing the right value of K | Introduction to decision trees | Splitting criteria: Gini and Entropy | Overfitting and pruning in decision trees | Comparing logistic, KNN, and decision trees

Live Class 2: What is a confusion matrix | Understanding TP, FP, TN, FN | Calculating precision and recall | What is F1 score and why it matters | Balancing precision and recall | Introduction to ROC curve | Interpreting ROC-AUC score | Choosing metrics based on problem type

Project: Titanic Dataset

Live Class 1: Introduction to neural networks | Structure of a neural network: neurons, layers, activation functions | Forward propagation process | Calculating outputs in a neural network | What is backpropagation | Adjusting weights during backpropagation | Gradient descent and learning rate | Role of loss function in training | Training a neural network through epochs

Live Class 2: Deep Neural Networks, Vanishing/Exploding Gradients, and Batch Normalization (with TensorFlow)

Project: MNIST Digit Classifier

Live Class 1: Object Detection Model Finetuning (Roboflow Journey) | Yolo Model Finetune

Live Class 2: Transformers : The core of Modern AI

Project: Sales Object Tracking & Heatmap Visualization