terminal — python3
>>> print("Python for Data Science & AI")
Python for Data Science & AI
>>> # A comprehensive research presentation
>>> author = "Prince Bhati" # Chandigarh University
>>> topics = [NumPy, Pandas, Matplotlib, Sklearn, TensorFlow]
>>> ▌
BTech AI & ML | Year 1 Foundations | Chandigarh University | 2025
W H Y P Y T H O N F O R D A TA S C I E N C E ?
The #1 Language for AI and Data
✦ Simple Syntax ◉ Versatile
Web, scripting, automation, APIs, ML, research — one language to rule them
Reads like pseudocode — minimal boilerplate, faster prototyping
all
◈ Rich Ecosystem ★ Industry Standard
Google, Meta, Netflix, OpenAI, Anthropic all use Python as their core ML
1000s of ML/DS libraries: NumPy, Pandas, Sklearn, TensorFlow, PyTorch
language
❋ Community ⬡ Open Source
Free, platform-independent, and continuously improved by thousands of
Largest DS/AI community — Stack Overflow, Kaggle, GitHub, arXiv
contributors
Python for Data Science & AI | Prince Bhati | Chandigarh University
T H E P Y T H O N D ATA S C I E N C E S TA C K
Essential Libraries Every Data Scientist Must Know
▸ N-dimensional arrays (ndarray)
NumPy import numpy as np
▸ Linear algebra operations a = [Link]([1,2,3])
A = [Link]((3,3))
▸ Random number generation
▸ DataFrame & Series objects
Pandas import pandas as pd
▸ Read CSV/Excel/JSON/SQL df = pd.read_csv('[Link]')
[Link]('col').mean()
▸ Data cleaning & reshaping
▸ Line, bar, scatter plots
Matplotlib import [Link] as plt
▸ Histograms, heatmaps, pie [Link](x, y)
[Link]('[Link]')
▸ Subplots & figure control
▸ Classification & Regression
Scikit-Learn from [Link] import
▸ Clustering (K-Means, DBSCAN) RandomForestClassifier
[Link](X_train, y_train)
▸ Dimensionality reduction (PCA)
▸ Deep learning framework
TensorFlow import tensorflow as tf
▸ Define & train neural networks model = [Link]()
[Link](optimizer='adam')
▸ Keras high-level API
Python for Data Science & AI | Prince Bhati | Chandigarh University
D ATA P R O C E S S I N G W O R K F L O W W I T H P Y T H O N
From Raw Data to ML-Ready Dataset
Load Data Split Data X_train, X_test, y_train, y_test
1 5
=
df = pd.read_csv('[Link]')
print([Link], [Link])
train_test_split(X,y,test_size=0.
2)
Explore (EDA) Train Model
2 6
[Link]() # Stats
model = RandomForestClassifier()
[Link]().sum() # Missing
[Link](X_train, y_train)
values
Clean Data Evaluate
3 7
[Link](inplace=True) from [Link] import
df['col'].fillna(df['col'].mean() accuracy_score,
) classification_report
Feature Eng. Deploy
4 8
import pickle
from [Link] import
[Link](model,
StandardScaler, LabelEncoder
open('[Link]','wb'))
Python for Data Science & AI | Prince Bhati | Chandigarh University
ML ALGORITHMS IMPLEMENTED IN PYTHON
Scikit-Learn Cheat Sheet
Regression Clustering
sklearn.linear_model.LinearRegress
Linear Regression ion
K-Means [Link]
Ridge/Lasso sklearn.linear_model.Ridge/Lasso DBSCAN [Link]
[Link]
SVR [Link] Agglomerative ering
Classification Dimensionality
sklearn.linear_model.LogisticRegre
Logistic Regression ssion
PCA [Link]
[Link]
Random Forest ifier
t-SNE [Link]
XGBoost [Link] LDA sklearn.discriminant_analysis.LDA
Python for Data Science & AI | Prince Bhati | Chandigarh University
D ATA V I S U A L I Z AT I O N I N P Y T H O N
Telling Stories With Data
Line Plot Scatter Plot Histogram
lib: matplotlib lib: seaborn lib: matplotlib
use: Time series, trends use: Correlation, clusters use: Distribution of data
[Link](x, y) [Link](data=df, [Link](df['col'],
[Link]('X'); [Link]('Y') x='a', y='b', hue='c') bins=30, color='steelblue')
Heatmap Box Plot Pair Plot
lib: seaborn lib: seaborn lib: seaborn
use: Correlation matrix use: Outliers, quartiles use: Multi-feature EDA
[Link]([Link](), [Link](x='cat', [Link](df,
annot=True, cmap='Blues') y='val', data=df) hue='target_class')
Python for Data Science & AI | Prince Bhati | Chandigarh University
PYTHON CAREER ROADMAP IN AI/ML
From Beginner to Industry-Ready
Beginner (0-3 Months) Intermediate (3-8 Months) Advanced (8-18 Months) Professional (18+ Months)
▸ Deep Learning with
▸ Python basics (loops, functions, OOP) ▸ Scikit-learn ML algorithms ▸ Production ML systems at scale
TensorFlow/PyTorch
▸ NumPy & Pandas fundamentals ▸ Feature engineering & EDA ▸ NLP with HuggingFace Transformers ▸ RAG systems, LLM fine-tuning
▸ Data visualization with Matplotlib ▸ Kaggle competitions (Bronze medals) ▸ MLOps: Docker, CI/CD, MLflow ▸ Research paper implementations
▸ Jupyter Notebooks, Git/GitHub ▸ Flask/FastAPI for model serving ▸ Deployed projects on GitHub + Cloud ▸ Targeting 20+ LPA roles (India)
💡 Pro Tip: Build 3 deployed projects + consistent GitHub commits + Kaggle profile = Internship/Job ready!