introduction-to-data-science-210911034830 (1).pdf
- 1. Introduction to Data Science DR. AHMAD KARAWASH APRIL 2021
- 2. Outline What is Data Science? Where does data come from? Why the excitement? Machine Learning Models The Data Science Process 9/10/2021 2
- 3. What is Data Science? History & definition 9/10/2021 3
- 4. Data Science Definition This is an Adaptation of the original Data Science Venn Diagram which is licensed under Creative Commons Attribution-NonCommercial. What is Data Science? Data Science is an art of analyzing and extracting knowledgeable information from the data. 9/10/2021 4
- 5. Where does data come from? 9/10/2021 5
- 6. Big Sources of Data Data-sources-of-Big-Data.ppm (640Ă360) (researchgate.net) Where does data come from? 9/10/2021 6
- 8. How Data Science Helped Southwest Save $100 Million on Fuel? How Big Data and the Industrial Internet Can Help Southwest Save $100 Million on Fuel | GE News Why the excitement? 9/10/2021 8
- 9. How Data Science helped UPS save 39 millions of Gallon by route optimization? big-data-infographic_v06 (ups.com) Why the excitement? 9/10/2021 9
- 10. Amazon Recommendation System 9/10/2021 10 Amazon - recommendation engine | Youssef Rahoui | Flickr
- 11. Netflix Movie Suggestion 9/10/2021 11 https://www.slideshare.net/xamat/qcon-sf-2013-machine-learning-recommender-systems- netflix-scale
- 12. US Polling: 2008 & 2012 Nate Silver uses a simple idea â taking a principled approach to aggregating polling instead of relying on punditry â and: ⢠Predicts 49/50 states in 2008 ⢠Predicts 50/50 states in 2012 https://hbr.org/2012/11/how-nate-silver-won-the-2012-p Why the excitement? 9/10/2021 12
- 14. Machine Learning models Machine Learning Models Supervised learning Predict an outcome based on historical patterns Ex: advertisements clicks Unsupervised learning Analyze the relationships between data elements Ex: Facebook âfriend suggestionâ Reinforcement learning Learns from mistakes and get better Ex: Deep Blue 9/10/2021 14
- 15. The Data Science Process 9/10/2021 15
- 16. The Data Science Process Formulate the question Generate Hypothesis Collect the data Clean Data Explore/Transform Data Build Machine Learning models Evaluate & Deploy the Model The Data Science Process 9/10/2021 16
- 17. Formulate the question ⌠Understand the business problem ⌠Excellent results depend on a better understanding of the problem. The Data Science Process 9/10/2021 17
- 18. Generate Hypothesis ⢠Guess an approach through which we derive some essential data parameters that have a significant correlation with the prediction target. ⢠Letâs take an example of loan approval prediction, some features: ⌠Income: If an applicant has a higher income, should get a loan easily. ⌠Education: Higher education results in higher income, so we can approve the loan request. ⌠Loan Amount: Lesser the amount, chances of loan approval are high. 9/10/2021 18
- 19. Collect the data Kaggle UCI DSE NCBI Kaggle: Your Machine Learning and Data Science Community UCI Machine Learning Repository Dataset Search (google.com) National Center for Biotechnology Information (nih.gov) The Data Science Process Gathering data from relevant sources! 9/10/2021 19
- 20. Clean Data Some common source of data errors: ⌠Duplicate entries ⌠Missing values 5 Data Science Projects That Will Get You Hired in 2018 - KDnuggets The Data Science Process 9/10/2021 20
- 21. Explore/ Transform Data Feature Identification Univariate Analysis Multi-variate Analysis Feature Engineering The Data Science Process 9/10/2021 21
- 22. Build Machine Learning models The Data Science Process Algorithm Selection Traing Model Model Prediction 9/10/2021 22
- 23. Evaluate & Deploy the Model Validate selected Machine learning model Deploy model and make it accessible through an API The Data Science Process 9/10/2021 23