MLClass started in October 2011
#HASHTAGUNSET on Twitter, identi.ca, TwitterStreamGraphs, flickr
Own objectives
- application to PIM
- pragmatic view of AI
Finished with 100% overall (cf signed letter).
A perfect score but more importantly a clearer view of the field. The score has to be distinguished from AIClass though since here it was possible to know your score and re-submit answers accordingly (within certain limits and timeness). As for AIClass though the main benefit is to enter a lifelong learning process (cf Education#InteractiveClasses).
Attended activities
- Introduction
- Linear regression with one variable
- (Optional) Linear algebra review
- Linear regression with multiple variables
- Octave tutorial
- Logistic Regression
- One-vs-all Classification
- Regularization
- Neural Networks
- Backpropagation Algorithm
- Practical advise for applying learning algorithms
- How to develop and debug learning algorithms
- Feature and model design, setting up experiments
- Support Vector Machines (SVMs)
- Survey of other algorithms: Naive Bayes, Decision Trees, Boosting
- Unsupervised learning: Agglomerative clustering, k-Means, PCA
- Combining unsupervised and supervised learning.
- (Optional) Idependent component analysis
- Anomaly detection
- Other applications: Recommender systems. Learning to rank
- Large-scale/parallel machine learning and big data.
- Machine learning design / practical methods
- Team design of machine learning systems
Overall remarks and conclusions
- meet him
- could be improved by doing that
- see also that other event
Back to the Menu
Other reviews or coverage
See also
To do
- see AIClass
- can artificial data synthesis used in OCR be considered a minimal form of artificial creativity?
Fabien Benetou's PIM
