Certificate Program in Advanced Python: From Analytics to AI

This week, you will learn how to extract value from raw organizational data. Explore the tools and libraries employed at every stage of the data lifecycle, spanning from collection to analysis, interpretation, and the generation of insightful information. These insights aim to guide decision-making and enhance business outcomes.

This week, you will delve into Python fundamentals, covering syntax, variables, data types, and operators. Explore key data structures such as lists, tuples, sets, and dictionaries, and master concepts like conditions and loops for effective program control flow. By the end of the week, you will be proficient in writing basic Python code, equipped with a solid understanding of its structure and fundamental constructs.

Below are the topics covered in this week:

• Python - why, what, how?
• Basic syntax, variables, data types, operators, expressions, and statements
• Lists, Tuples, Sets, Dictionaries
• Conditions
• Loops

This week, you will learn the principles of Object-Oriented Programming (OOPS) to design and implement modular, reusable code. Acquire proficiency in creating and utilizing functions for effective code organization, master file handling for data manipulation, explore the integration of APIs for external data retrieval, and refine your ability to handle exceptions for robust error management in Python programs.

Below are the topics covered in this week:

• OOPS
• Functions
• Modules

This week, you will learn Exploratory Data Analysis (EDA) techniques using Matplotlib, Seaborn, and Plotly for data visualization. Gain a deep understanding of data distributions and correlations and develop skills in hypothesis testing. This will empower you to draw meaningful conclusions and make informed decisions through statistical analysis of diverse datasets.

Below are the topics covered in this week:

• EDA
• Matplotlib
• Seaborn
• Data Distributions
• Correlations

 

This week, you will learn to load and read data proficiently using Pandas and perform data processing, cleaning, and transformation on raw datasets. Enhance your skills in manipulating and reshaping data, facilitating thorough analysis and preparation for further exploration and modeling in Python.

Below are the topics covered in this week:

• Load and read data using pandas
• Data Processing
• Data Cleaning
• Data Transformations
• Data Manipulation

Technoecom, a leading Australian retailer, offers over 500 high-quality electronic products and recently launched an e-commerce platform to enhance its digital presence. However, during a three-month pilot, they faced challenges with payment methods, order volume, geographical coverage, shipping, and customer satisfaction. Now, they seek to evaluate the feasibility of expanding their digital operations to more cities and assess pilot program revenue. As the analytics lead, you must analyze the data to provide insights for the Chief Experience Officer's decision on the expansion

This week covers the introduction to Hugging Face and its features and getting setup with creating spaces along with secret keys.

The outcome of this week is to comprehend the principles of supervised and unsupervised learning. Gain proficiency in train and evaluating model performance using metrics, fine-tuning models for optimal results, and conducting comparative analyses to choose the most effective machine learning approach for diverse tasks.

Below are the topics covered in this week:

•  Classification
• Regression
• Metrics - Scores
• Tuning Models
• Comparative analysis of models

This week, you will learn how to deploy machine learning models, focusing on making inferences and predictions in real-world applications. Acquire practical experience with Gradio, a user-friendly library for creating interfaces to interact with models, enhancing your capability to deploy and effectively showcase machine learning solutions.

Below are the topics covered in this week:

• Model Deployment
• Inferences
• Gradio
• Performance Reporting

This week, you will learn to use Python to establish resilient MLOps pipelines, covering model versioning, continuous integration, and deployment automation. Develop the capability to streamline the entire machine learning lifecycle, fostering collaboration between data scientists and operations teams. By the week's end, learners should be adept at constructing scalable and maintainable machine-learning systems in a production environment.

HealthyLife, based in NYC, faces challenges in accurately predicting insurance charges using traditional methods. The company struggles with underpricing and overpricing, impacting profitability and customer satisfaction. To address this, they seek to implement advanced predictive modelling for personalized and accurate pricing. The objective is to improve charge prediction accuracy, streamline the underwriting process, and maintain regulatory compliance. Goals include fair pricing, reduced risk of pricing errors, enhanced customer satisfaction, and strengthened market position through better transparency and compliance.

This week covers the Project Q&A + AI Refresher - Introduction to Generative AI
 

The learning for this week includes an exploration of Prompt Engineering, Anyscale API, and at the end of the week, practical exposure to building Generative AI applications on classification and summarization tasks through a hands-on MLS.

Below are the topics covered in this week:

• Generative AI
• ChatGPT
• LLM
• Prompt
• Code generation
• Data generation

This week, you will learn the basics of vector databases, gaining an introduction to the fundamentals of vector data and its structure. Explore vector embeddings and search techniques, empowering you to execute efficient queries on vector databases for applications like similarity search and recommendation systems.

Below are the topics covered in this week:

• Introduction to vector databases
• Understanding vector data
• Vector embeddings and vector search
• Querying on vector database

The outcome of this week is to learn how to create an RAG solution that can answer questions based on pre-existing documents. Learn how to host your RAG on HuggingFace Spaces for inference. At the end of the week, we will deploy a Gradio app on Hugging Face which can answer questions about the streamlined documentation replicating the Langchain Chatbot.

• Introduction to RAG
• RAG Workflow (Indexing, Retrieval and Augmentation)
• HuggingFace Spaces
• RAG Deployment and Logging

Finsights Grey Inc., a financial tech firm, struggles with the manual extraction of insights from extensive 10-K reports, slowing down their recommendation process. They aim to implement a Retrieval-Augmented Generation (RAG) model to automate this process and enhance the accuracy and speed of their analysts. The goal is to develop a Gradio app on HuggingFace spaces for efficient information extraction and analysis, focusing on 10-k reports of major cloud and AI Organisations like AWS, Google Cloud, Azure, Meta AI, and IBM. Success will be measured by the system’s ability to accurately answer specific questions about these companies and provide sources of the LLM’s answer, improving analysts' productivity, providing timely insights, and strengthening the company's market position.