Introduction to Data Science

01 Definition

What is data science?

Data science is the practice of collecting, cleaning, analyzing, and interpreting data to answer questions and support decisions. A data scientist uses statistics to find patterns, programming to process large datasets, machine learning to build predictive models, and visualization to explain the results to other people.

A complete overview of data science

Data science compared with related fields

02 How it works

The data science lifecycle

Most projects follow the same loop. You will often go back a step when you learn something new.

1. AskDefine the question, the goal, and how success is measured.
2. CollectGather data from databases, files, APIs, sensors, or surveys.
3. CleanFix missing values, duplicates, wrong types, and outliers.
4. ExploreUse statistics and charts to understand patterns and problems.
5. ModelBuild and train a statistical or machine learning model.
6. EvaluateTest on unseen data and compare with a simple baseline.
7. CommunicatePresent results with clear visuals and honest limits.
8. Deploy and monitorPut it into use and track performance over time.

Where the time really goes

This split is only an illustration. Real projects vary, but data preparation usually takes the biggest share.

03 Types and roles

Types of data science work

Four levels of analytics

Types of data

Common job roles

04 Core concepts

What you need to learn

Key terms

Dataset

A collection of data, usually rows (records) and columns (features).

Feature

An input column used to describe each record, such as age or price.

Target

The value you want to predict, such as whether a customer churns.

Correlation

A relationship between two variables. It does not prove that one causes the other.

Outlier

A value that is very different from the rest. It may be an error or an important signal.

Overfitting

A model that memorizes training data and performs badly on new data.

Cross validation

Testing a model on several splits of the data to get a more reliable score.

A/B test

A controlled experiment comparing two versions to see which performs better.

05 How to build

How to build a data science project

1. Choose a clear question

   Example: “Which customers are likely to cancel in the next 30 days?” A narrow question makes the whole project easier.

2. Find data

   Use company data, public sources like Kaggle datasets and government portals, or collect your own with APIs or surveys.

3. Clean the data

   Check data types, fill or remove missing values, remove duplicates, and investigate outliers.

4. Explore it

   Plot distributions and relationships. Look for surprises and write down what you learn.

5. Build a baseline, then a model

   Start simple, such as an average or logistic regression. Only move to complex models if they clearly help.

6. Evaluate on unseen data

   Use the right metric, such as accuracy, recall, or error size, and compare with the baseline.

7. Present and publish

   Create a clear dashboard or short report. Put your code and findings on GitHub.

import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression

df = pd.read_csv("customers.csv")      # load data
df = df.dropna()                       # clean missing values

X = df[["age", "monthly_usage", "support_calls"]]
y = df["churned"]                      # what we want to predict

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)

model = LogisticRegression()
model.fit(X_train, y_train)            # training
print(model.score(X_test, y_test))     # evaluation

06 Tool stack

Tools used in data science

Tools and features change often, so check official websites before choosing.

07 Real world uses

Where data science is used

Portfolio project ideas for beginners

08 Data science and AI

How data science connects to AI

Data science and AI overlap heavily. Machine learning is the bridge: it is a tool inside data science and the core engine of most modern AI.

09 Pitfalls and ethics

Common mistakes and responsible use

10 Career roadmap

Your learning path

1. Month 1Excel, basic statistics, and SQL queries.
2. Month 2Python, pandas, and NumPy. Clean real datasets.
3. Month 3Visualization, EDA, and a first dashboard.
4. Month 4Machine learning with sklearn. Build two models.
5. Month 5One bigger project, Git, and a portfolio page.
6. Month 6Apply for roles, practice interviews, and learn deployment.

11 FAQ

Frequently asked questions