Why We Start Data Science with Prompt Engineering

Описание: In this video, Professor Nik Bear Brown outlines the first third of the course. Before we dive into the heavy lifting of Exploratory Data Analysis (EDA), Statistical Hypothesis Testing, and Feature Selection, we are taking a "strategic detour" into Prompt Engineering.

Why? Because in 2026, a productive Data Scientist uses Language Models to accelerate the writing of cleaning and visualization scripts. We will learn how to take existing scripts from the class GitHub and use LLMs to audit, improve, and adapt them to new data challenges.

Key Topics Covered:

The Immutable Raw: Why we always keep df_raw and the importance of tracking data transformations.

The Detour: Transitioning from "how to compute" to "how to prompt the computer."

Module 1 Scope: Data Pre-processing, Visualization, and a review of core Machine Learning concepts.

GitHub as a Base: How to use the thousands of existing class examples as a starting point for LLM-assisted refinement.

Module 1 Roadmap:

Introduction to the Data Pipeline

Special Detour: Prompting for Technical Scripts

Understanding Data Distributions & Statistics

Data Integration and Pre-processing Techniques

EDA and Visualization Mastery

Resources:

GitHub Repository: https://github.com/nikbearbrown/INFO_...

#DataScience #INFO7390 #PromptEngineering #DataPreprocessing #EDA #PythonDataScience #GenerativeAI #LLM #NortheasternUniversity #NikBearBrown #DataVisualization