UNIVERSITY OF ECONOMICS HO CHI MINH CITY
SCHOOL OF ECONOMIC MATHEMATICS AND STATISTICS
INTRODUCTION TO DATA SCIENCE AND
APPLICATIONS
Vietnam Netherlands Programme - 2019
Instructor: TRAN THI TUAN ANH
Chapter 4: Real-world Applications of Machine Learning
Chapter 4: Real-world Applications of Machine
Learning
Instructor: TRAN THI TUAN ANH 2 / 36
Chapter 4: Real-world Applications of Machine Learning
Chapter’s outline
1 Data Science and Business Strategy
2 Some real-world applications of machine learning
3 Solving some specific tasks with machine learning algorithm
Instructor: TRAN THI TUAN ANH 3 / 36
Chapter 4: Real-world Applications of Machine Learning
1. Data Science and Business Strategy
Discussion: how to apply data science in real-life scenarios in Vietnam?
Instructor: TRAN THI TUAN ANH 4 / 36
Chapter 4: Real-world Applications of Machine Learning
1. Data Science and Business Strategy
Discussion: how to apply data science in real-life scenarios in Vietnam?
Healthcare Analytics
Agriculture and Farming
Finance and Banking
E-commerce
Transportation and Logistics
Smart Cities
Education
Customer Experience and Feedback
Environmental Sciences
Instructor: TRAN THI TUAN ANH 5 / 36
Chapter 4: Real-world Applications of Machine Learning
1. Data Science and Business Strategy
Instructor: TRAN THI TUAN ANH 6 / 36
Chapter 4: Real-world Applications of Machine Learning
1. Data Science and Business Strategy
Data-Analytic Thinking and Data-Driven Decision Making
The firm’s management must think data-analytically;
The management must create a culture where data science, and data
scientists, will thrive.
Achieving and sustaining competitive advantages with data science
Attracting data scientists and their teams
Instructor: TRAN THI TUAN ANH 7 / 36
Chapter 4: Real-world Applications of Machine Learning
1. Data Science and Business Strategy
Be Ready to Accept Creative Ideas from Any Source
Once different role players understand fundamental principles of data
science, creative ideas for new solutions can come from any direction
Be Ready to Evaluate Proposals for Data Science Projects
Is the business problem well specified? Does the data science solution
solve the problem?
Is it clear how we would evaluate a solution?
Would we be able see evidence of success before making a huge
investment in deployment?
Does the firm have the data assets it needs? For example, for
supervised modeling, are there actually labeled training data? Is the
firm ready to invest in the assets it does not have yet?
Instructor: TRAN THI TUAN ANH 8 / 36
Chapter 4: Real-world Applications of Machine Learning
2. Machine Learning Applications in Real World
Some of the best machine learning examples
Traffic Alerts
Image Recognition
Video Surveillance
Sentiment Analysis
Product Recommendation
Online support using Chatbots
Google Translate
Machine Learning Usage in Social Media
... and many other applications
Instructor: TRAN THI TUAN ANH 9 / 36
Chapter 4: Real-world Applications of Machine Learning
2. Machine Learning Applications in Real World
2.1. Traffic Alerts
How does Google Maps know that you are on the fastest route despite the
traffic is high?
Data is captured and stored by the app "Google Maps"
Your location
Your average traveling speed
Answers to the questions like "does the route still have traffic"?
Day, time, and any specific occasion
By using this data, AI & machine learning algorithms make the right
conclusions and give you the exact information.
The updated feature of Google Maps also helps us to know how far is
the upcoming bus from a specific stop; make predictions on the bus
delays; how crowded the TRAN
Instructor: bus THI
or TUAN
[Link] 10 / 36
Chapter 4: Real-world Applications of Machine Learning
2. Machine Learning Applications in Real World
2.2. Image Recognition
So how does machine learning go about facial recognition?
Phone unlocking
Drones - unmanned, aerial devices
Self-driving cars
Forest Activities
...
Instructor: TRAN THI TUAN ANH 11 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
Predicts and suggests relevant items to users based on their
preferences and interests
Widely used in various applications and industries, including
e-commerce, entertainment, social media, and content streaming
platforms.
here are primarily two types of recommendation systems:
Content-Based Filtering: recommends items to users based on their
previous actions and profile.
Collaborative Filtering: recommends items to users based on the
preferences and behaviors of similar users.
Hybrid Recommendation Systems
Instructor: TRAN THI TUAN ANH 12 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
Instructor: TRAN THI TUAN ANH 13 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
Instructor: TRAN THI TUAN ANH 14 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
KNN Recommender System for Movie Recommendation -
Collaborative Filtering
Instructor: TRAN THI TUAN ANH 15 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
KNN Recommender System for Movie Recommendation :
Collaborative Filtering - KNNBasic
KNNBasic stands for K-Nearest Neighbors Basic.
It finds people who have similar tastes to yours.
It looks at what those people have liked that you haven’t seen yet.
Based on what those similar people have liked, it predicts what you
might like.
Instructor: TRAN THI TUAN ANH 16 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
KNN Recommender System for Movie Recommendation :
Collaborative Filtering - KNNBaseline
It finds people who have similar tastes, just like the basic version.
Calculating baseline ratings for each person and each item
Calculate the Overall Mean Rating across all users and items. and
adjust user and item Biases
Identify the functional relationships between the input (feature) and
the desired output (label) to adjust the biases.
Applying KNN algorithm and predict ratings
Instructor: TRAN THI TUAN ANH 17 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
KNN Recommender System for Movie Recommendation :
Collaborative Filtering - KNNWithMeans
It finds people who have similar tastes to yours.
It calculates the average rating for each user and each item.
It then adjusts the ratings by subtracting these average ratings from
the raw ratings.
After this adjustment, it uses the adjusted ratings to predict what you
might like.
Instructor: TRAN THI TUAN ANH 18 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
KNN Recommender System for Movie Recommendation :
Content-Based Filtering
Calculate Similarities by using a similarity measure to identify
K-Nearest Neighbors:
Recommend these neighbor items to users
Incorporate methods such as feature extraction, dimensionality
reduction, or hybrid approaches to address the cold start problem
effectively.
Instructor: TRAN THI TUAN ANH 19 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
KNN Recommender System for Movie Recommendation :
Content-Based Filtering
Instructor: TRAN THI TUAN ANH 20 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
Other algorithms for Recommendation systems
Decision Tree
K-Means
Associate rules
Support Vector Machine
Instructor: TRAN THI TUAN ANH 21 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
Recommendation systems by Decision Tree
Instructor: TRAN THI TUAN ANH 22 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
Recommendation systems by Decision Tree
Instructor: TRAN THI TUAN ANH 23 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
Recommendation systems by clustering with K-Means
Instructor: TRAN THI TUAN ANH 24 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
Recommendation systems by associate rules
Gather transactional data and organize the data into a format
suitable for association rule mining.
Apply an association rule mining algorithm to determine the itemsets
that occur together frequently
Select rules that meet the desired thresholds for confidence and lift,
indicating strong associations between items.
Instructor: TRAN THI TUAN ANH 25 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
Recommendation systems by associate rules
Instructor: TRAN THI TUAN ANH 26 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.1. Recommendation systems
Recommendation System by combining many ML algorithms
Instructor: TRAN THI TUAN ANH 27 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.2. What kinds of issues can be solved with Decision Tree?
Predicting whether an email is spam or not spam.
Customer churn prediction
Market segmentation
Credit scoring and risk assessment
Product recommendation and cross-selling
Fraud detection
Employee performance evaluation
Market analysis and customer insights
...
Instructor: TRAN THI TUAN ANH 28 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.2. What kinds of issues can be solved with Decision Tree?
Decision tree for Customer churn prediction
Instructor: TRAN THI TUAN ANH 29 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.2. What kinds of issues can be solved with Decision Tree?
Decision tree for Market segmentation
Instructor: TRAN THI TUAN ANH 30 / 36
Chapter 4: Real-world Applications of Machine Learning
3. Solving some specific tasks with machine learning
algorithm
3.2. What kinds of issues can be solved with Decision Tree?
Decision tree for fraud detection
Instructor: TRAN THI TUAN ANH 31 / 36
Chapter 4: Real-world Applications of Machine Learning
2. Machine Learning Applications in Real World
2.3. Video Surveillance
Surveillance cameras are video cameras used for the purpose of
observing an area.
The Video surveillance system is composed of 5 parts: camera,
transmission, control, display, record registration.
Video surveillance is used for different purposes like:
Copper theft prevention
Abnormal event detection
Facility protections
Parking lots
Traffic monitoring
...
Instructor: TRAN THI TUAN ANH 32 / 36
Chapter 4: Real-world Applications of Machine Learning
2. Machine Learning Applications in Real World
2.4. Sentiment Analysis
Sentiment analysis is a top-notch machine learning application that
refers to sentiment classification, opinion mining, and analyzing
emotions.
Using this model, machines groom themselves to analyze sentiments
based on the words. They can identify if the words are said in a
positive, negative, or neutral notion.
This applies Natural Language Processing (NLP) techniques
Also, they can define the magnitude of these words.
Instructor: TRAN THI TUAN ANH 33 / 36
Chapter 4: Real-world Applications of Machine Learning
2. Machine Learning Applications in Real World
2.5. Online support using Chatbots
When you use any application, you see an option called "chat with
us".
These are chatbots running on the concepts of machine learning.
Chatbots can be trained to respond customer service questions and
provide fast solutions, 24/7.
These bots can recognize the type of questions, and accordingly, give
quick answers to resolve the query by extracting the right parameters.
Any errors they make can be quickly identified and corrected to
ensure the chatbot doesn’t repeat them in the future.
For businesses, customer service chatbots offer a way to reduce costs
and satisfy customers by providing fast and convenient service.
Instructor: TRAN THI TUAN ANH 34 / 36
Chapter 4: Real-world Applications of Machine Learning
4. Group discussion
Group discussion to prepare for final project
Topic of final project
Outline of the content
Assign tasks for each member ]itemSchedule for preparation
Instructor: TRAN THI TUAN ANH 35 / 36
Chapter 4: Real-world Applications of Machine Learning
THE END
THANK YOU FOR LISTENING
Instructor: TRAN THI TUAN ANH 36 / 36