Scribd
Upload
18 views22 pages

Class 2 - Lifecycle ML Concepts in Ds

The document outlines the Data Science Life Cycle, detailing its phases such as problem formulation, data extraction, preprocessing, modeling, and deriving actionable insights. It compares Data Science with Business Intelligence and Data Mining, highlighting their distinct focuses and methodologies. Additionally, it discusses various applications of Data Science across industries, emphasizing the role of machine learning algorithms in solving real-world problems.

Uploaded by

techboys2022
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
18 views22 pages

Class 2 - Lifecycle ML Concepts in Ds

The document outlines the Data Science Life Cycle, detailing its phases such as problem formulation, data extraction, preprocessing, modeling, and deriving actionable insights. It compares Data Science with Business Intelligence and Data Mining, highlighting their distinct focuses and methodologies. Additionally, it discusses various applications of Data Science across industries, emphasizing the role of machine learning algorithms in solving real-world problems.

Uploaded by

techboys2022
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 22

CLASS 2

• Data science Life Cycle


• Applications
• Business Intelligence Vs Data Science
• Data Mining Vs Data Science
Data Science Life Cycle
Data Science Life Cycle

• Definition: Data Science Life Cycle defines the process of the how
information is carried out in various phases for professionals
working on a project.

• It’s a step-by-step procedure that is arranged in a circular


structure.

• Each phase has its own characteristics and importance


Data Science Life Cycle
Formulating a Business Problem
• Definition: Identify a business challenge that can be addressed
using data science techniques.

• Importance: Ensures that the project is aligned with business


objectives and has a clear goal.

• Example: A retail store wants to predict sales for the next quarter
to optimize inventory management and reduce waste of perishable
products. This involves analyzing historical sales data and external
factors like seasonal trends or economic conditions.
Data Extraction, Transformation, and
Loading (ETL)
• Definition: Extract relevant data from various sources, transform
it into a machine-readable format, and load it into the system for
analysis.

• Importance: Ensures that the data is consistent, complete, and


ready for modeling.

• Example: For the retail store example, gather past sales data,
weather patterns, holidays, and any promotional activities.
Transform this data into a structured format suitable for analysis.
Data Preprocessing
• Definition: Clean, manipulate, and prepare the data for modeling
using techniques like exploratory data analysis (EDA), statistical
analysis, and data wrangling.

• Importance: Improves data quality and helps identify patterns or


correlations.

• Example: Convert sales data into a time-series format, check for


stationarity, and identify trends or seasonality. Remove outliers or
handle missing values to ensure data integrity.
Data Modeling
• Definition: Apply machine learning or statistical models to the
preprocessed data to solve the business problem.

• Importance: Enables the extraction of insights from complex data.

• Example: Use time-series forecasting models like ARIMA, SARIMA,


or Prophet to predict future sales based on historical patterns and
trends.
Gathering Actionable Insights
• Definition: Interpret the outputs from the models to derive
meaningful insights that address the business problem.

• Importance: Translates technical results into business language.

• Example: Analyze the forecasted sales trends to understand


expected demand for different products over the next quarter.
Solutions for the Business Problem
• Definition: Develop actionable recommendations based on the
insights gathered.

• Importance: Ensures that the insights lead to tangible business


outcomes.

• Example: Use the sales forecasts to adjust inventory levels,


prioritize stock replenishment for fast-moving items, and
implement strategies to minimize waste of perishable goods.
Applications / Benefits of Data Science
Applications of Data Science Across
Industries
Banking and Finance

• Fraud detection using machine learning models (e.g., supervised and unsupervised techniques).
• Personalized financial advice through customer segmentation and sentiment analysis.

Retail

• Enhance customer experience and retention.


• Example: Personalized product recommendations based on purchase history and browsing behavior.

Healthcare

• Analyze data from wearable trackers to monitor patient health.


• Optimize hospital operations by reducing waiting times and improving care quality.
Applications in Other Key Sectors
Transportation
Improve customer journeys with personalized travel details.
Example: Transport for London maps customer journeys and manages disruptions using statistical data.

Construction
Track task completion times, material costs, and resource usage for better decision-making.

Manufacturing
Analyze untapped data from manufacturing processes to optimize efficiency and reduce waste.

Energy and Utilities


Study utility consumption patterns to enhance control and improve consumer feedback.
Advanced Applications of Data Science
Geospatial Analysis
• Analyze massive graphical, temporal, and geospatial data.
• Applications include seismic interpretation and reservoir characterization.

Media and Entertainment


• Leverage social media content for real-time usage pattern analysis.
• Create audience-specific content, measure performance, and recommend on-demand media.

Public Services
• Use cases include health research, fraud detection, energy exploration, environmental protection,
and financial market analysis.
Business Intelligence Vs Data Science
Criterion Business Intelligence Data science

Data science deals with structured and


Data Business intelligence deals with structured data,
Source unstructured data, e.g., weblogs, feedback, etc.
e.g., data warehouse.

Method Analytical(historical data) Scientific(goes deeper to know the reason for


the data report)

Statistics and Visualization are the two skills Statistics, Visualization, and Machine learning are
Skills required for business intelligence. the required skills for data science.

Data science focuses on past data, present data,


Business intelligence focuses on both
Focus and also future predictions.
Past and present data
Data Mining Vs Data Science
Machine Learning in Data Science
Key Algorithms and Their Applications
Definition: Machine learning is a subset of artificial intelligence that
enables systems to learn and make decisions without explicit programming.

• Role in Data Science:


• Automates pattern recognition and prediction.
• Enhances data analysis with advanced algorithms.

• Types of Machine Learning:


• Supervised Learning
• Unsupervised Learning
• Reinforcement Learning
Importance of ML Algorithms in Data Science
Why Learn ML Algorithms?
• Solve real-world problems like predictions, classifications, and anomaly detection.
• Extract actionable insights from data.

Applications:
• Forecasting sales or trends.
• Customer segmentation.
• Fraud detection.

Key ML Algorithms
• Regression
• Decision Tree
• Clustering
• Classification
• Outlier Analysis
Key Algorithms and Their Applications
Definition: Machine learning is a subset of artificial intelligence that
enables systems to learn and make decisions without explicit programming.

• Role in Data Science:


• Automates pattern recognition and prediction.
• Enhances data analysis with advanced algorithms.

• Types of Machine Learning:


• Supervised Learning
• Unsupervised Learning
• Reinforcement Learning
Supervised/Unsupervised/Reinforcement
Feature Supervised Learning Unsupervised Learning Reinforcement Learning

Learns from unlabeled data by finding Learns through trial and error by
Definition Learns from labeled data (input-output pairs).
patterns. interacting with the environment.

Unlabeled data (no predefined No predefined data, learns from


Data Type Labeled data (with correct answers).
categories). experience.

Predict an outcome (classification or Discover hidden patterns or groupings in Maximize rewards through continuous
Goal
regression). data. interactions.

Linear Regression, Decision Trees, Neural K-Means Clustering, Principal Component Q-Learning, Deep Q-Networks (DQN),
Example Algorithms
Networks. Analysis (PCA), Autoencoders. Policy Gradient.

Spam detection, medical diagnosis, price Customer segmentation, anomaly Robotics, game playing (AlphaGo), self-
Use Cases
prediction. detection, recommendation systems. driving cars.

No explicit feedback; finds structure on its


Feedback Type Learns from correct output labels. Learns through rewards and penalties
own.

You might also like