* Start Your Journey toward Machine Learning!
Due to popular demand, Yidu AI, the machine learning community, will kick-start the 3rd part of the 4-week data science program on Nov 2nd. The course will include from basic understanding about Machine learning and sklearn(scikit learn) to Model development.
There are so many Machine Learning courses available, why choose Yidu AI?
- We have been in the education sector for 3 years and understand your pain point – let’s just start to code and see real results without spending too much time on the unnecessary details (however, we will still provide them for you to read up)
- Industry collaboration: come participate in our events with the industry leaders to bring to the Yidu AI community. Here is the most recent example. (https://www.meetup.com/Yidu-AI-Meetup/events/264860151/)
Register here: https://www.eventbrite.sg/e/machine-learning-with-scikit-learn-hands-on-tickets-74533077441
* Why learn Scikit learn?
Scikit-learn is a machine-learning library. Its goal is to provide a set of common algorithms to Python users through a consistent interface. This means that hard choices have to be made as to what fits into the project. For access to high-quality, easy-to-use, implementations1 of popular algorithms, scikit-learn is a great place to start
| Covers most machine-learning tasks |
Scan the list of things available in scikit-learn and you quickly realize that it includes tools for many of the standard machine-learning tasks (such as clustering, classification, regression, etc.). And since scikit-learn is developed by a large community of developers and machine-learning experts, promising new techniques tend to be included in fairly short order.
| Variety of Use Cases and Wide Adoption in the Industry |
- Image classification domain, Sklearn’s implementation of K-Means along with PCA has been used for handwritten digit classification very successfully in the past.
Sklearn has also been used for facial/ faces recognition using SVM with PCA. Image segmentation tasks such as detecting Red Blood Corpuscles or segmenting the popular Lena image into sections can be done using sklearn.
- Recommendation: A lot of us here use Spotify or Netflix and are awestruck by their recommendations. Recommendation engines started off with the collaborative filtering algorithm. To find out users with similar tastes, a KNN algorithm can be used which is available in sklearn. You can find a good demonstration of how it is used for music recommendation here.
- Classical data modeling can be bolstered using sklearn. Most people generally start their kaggle competitive journeys with the titanic challenge. It uses the robust Logistic Regression, Random Forest and the Ensembling modules to guide the user. You will be able to experience the user-friendliness of sklearn first hand. Sklearn has made machine learning literally a matter of importing a package.
- Sklearn also helps in Anomaly detection for highly imbalanced datasets (99.9% to 0.1% in credit card fraud detection) through a host of tools like EllipticEnvelope and OneClassSVM. In this regard, the recently merged IsolationForest algorithm especially works well in higher dimensional sets and has very high performance.
* Course Format and Hands-on exercise
| 4 Weekend courses |
1.5 hours teaching + 1.5 hours hands-on in-class project
| Hands-on Exercises |
We will prepare in-class project every week
Students will have opportunities to work with assistance from TAs
| Course Project |
Using sklearn to implement, deploy and maintain a practical credit scoring system
* About the Instructor
| Dr. Ouyang Ruofei |
A senior data scientist. He works on credit risk modeling in consumer lending business which serves millions of customers from Indonesia, India, and Vietnam. Prior to this position, he is a research fellow in Risk Management Institute of NUS focusing on financial derivative pricing model using machine learning.