Applications of Machine Learning
-
Upload
hayim-makabee -
Category
Data & Analytics
-
view
716 -
download
0
Transcript of Applications of Machine Learning
![Page 1: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/1.jpg)
Applications of Machine Learning
Hayim Makabee
July/2015
Predictive Analytics Expert
![Page 2: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/2.jpg)
2
Main applications of Machine Learning, by type of problem:
1. Clustering
2. Classification
3. Recommendation
Applications of Machine Learning
![Page 3: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/3.jpg)
3
Goal: Cluster observations into meaningful groups.
Clustering
![Page 4: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/4.jpg)
4
• “How to group customers for targeted marketing purposes?”
• “Which neighborhoods in a country are most similar to each other?”
• “What groups of insurance policy holders have high claim costs?”
• “How to group the products in a store based on their attributes?”
• “How to group pictures based on their description?”
Examples of Clustering Applications
![Page 5: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/5.jpg)
5
Goal: Predict class from observations.
Classification
![Page 6: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/6.jpg)
6
• “Which category of products is most interesting to this customer?”
• “Is this movie a romantic comedy, documentary, or thriller?”
• “Is this review written by a customer or a robot?”
• “Will the customer buy this product?”
• “Is this email spam or not spam?”
Examples of Classification Applications
![Page 7: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/7.jpg)
7
Goal: Personalized recommendation of items to users.
Recommendation
![Page 8: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/8.jpg)
8
• “Which movies should be recommended to a user?”
• “If the user just listened to a song, which song would he like now?”
• “Which news articles are relevant for a user in a particular context?”
• “Which advertisements should be displayed for a user on a mobile app?”
• “Which products are frequently bought together?”
Examples of Recommendation Applications
![Page 9: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/9.jpg)
9
In May 2015 Flickr released an automatic image tagging capability that mistakenly labeled a black man for an ape.
Soon afterwards, Google came up with a photo labeling tool similar to Flickr, which made similar mistakes. Black men were tagged as gorillas.
A recent Carnegie Mellon University study showed that Google displayed ads in a way that discriminated based on the gender of the user.
Failures of Machine Learning
![Page 10: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/10.jpg)
10
ML Process
Feature Engineering
Learning(Training)
Evaluation(Metrics)
Deployment(Serving)
![Page 11: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/11.jpg)
11
Derive new features from the initial data set:
1. Aggregations: Count, Sum, Min, Max, Avg, Std
2. Temporal: Elapsed time, Trends
3. Continuous to Categorical: Converting real values to enumerations.
4. Categorical to Binary: Converting enumerations to binary features.
5. Domain-specific derived features.
Feature Engineering
![Page 12: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/12.jpg)
12
Learning
![Page 13: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/13.jpg)
13
Question: How to evaluate the predictive accuracy of your model?
Answer: Partition the data set into Train and Test sets.
Train is like the “past” you learn from, and Test is like the “future” you predict.
Evaluation
![Page 14: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/14.jpg)
14
Evaluation of Binary Classification
![Page 15: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/15.jpg)
15
Binary Classification: Selecting the Threshold
![Page 16: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/16.jpg)
16
Accuracy = (True Positive + True Negative) / Total Population
Precision = True Positive / (True Positive + False Positive)
Recall = True Positive / ( True Positive + False Negative)
Normally there is a trade-off between Precision & Recall.
Business decision: Precision vs. Recall.
Metrics: Accuracy, Precision and Recall
![Page 17: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/17.jpg)
17
Questions? Comments? Concerns?
![Page 18: Applications of Machine Learning](https://reader036.fdocuments.net/reader036/viewer/2022062503/588285ec1a28ab24788b77bd/html5/thumbnails/18.jpg)
ThankYou!