Tags:
create new tag
view all tags

Machine Learning - A.Y. 2016/2017

Machine learning explores the study and construction of algorithms that can learn from and make predictions on data. Such algorithms operate by building a model from example inputs in order to make data-driven predictions or decisions, rather than following strictly static program instructions. Machine learning algorithms can be applied to virtually any scientific and non-scientific field (health, security and cyber-security, management, finance, automation, robotics, marketing..)

Instructor Telephone Office hours Studio
Paola Velardi 06-49918356 send e-mail Via Salaria 113 - 3° floor n. 3412

Course schedule

FIRST semester:

When   Where
Monday 10.00-14.00 aula alfa (ground floor)

Important Notes

The course is taught in English. Attending classes is HIGHLY recommended (homeworks, mid-term, laboratory)

Homeworks and self-assessment tests are distributed via the google group, you MUST register

Summary of Course Topics

The course introduces motivations, paradigms and applications of machine learning.

Topics Supervised learning: decision trees, instance-based learning, naīve Bayes, support vector machine, neural networks, deep learning, ensamble methods. Unsupervised learning: clustering. Semi-supervised learning. Reinforcement learning. Genetic algorithms and genetic programming. Issues in machine learning: underfitting, overfitting, model selection, error analysis.

Laboratory

Students will experiment the main algorithms using the Weka ML software Weka. Part of the experimental work is done in class, therefore students must install Weka and bring their labtops in class. Except for the first lesson, any other lesson consists in the presentation of a ML algorithm and the subsequent experimentation on WEKA. Detailed description of the experiments in each lab will be handled every week during the course.

Textbooks

Additional useful texts:

Resources:

Exam

  • Written exam on course material (60% of final grade)
  • 6-8 homeworks using the Weka ML toolkit
  • Weka project (40%)
  • rounding on final grade: +0.9 (max!!) depending on handled homeworks and presence in class
  • Examples of mid-term pdf pdf

Project 2016 (Fall): Predicting forest cover type

The project is described here

The data set can be downloaded from DRIVE https://drive.google.com/file/d/0By8kOQZC1_qZTXJLdGx0ZkdiX2c/view?usp=sharing

Please READ CAREFULLY the project description to understand what is expected and how the project is evaluated

Project 2016 (Spring): GAME WINNER PREDICTION CONTEST

The spring 2016 project was be a competion among student teams (max 3 students per team). The task is to predict the winner of a Role Playing Game (RPG) with direct clash. Students will be given a large dataset with detailed information on thousands of games, including the ID of the two competitors, the date of the match and the winner ID. The students will deliver the Predictor by the end of June (according to precise project specifications). Instructors will feed the systems with the details of additional games (not in the learning set) and compute the precision of each system at predicting the winner ID.

The project description is found here The learning dataset coun be downloaded here

Weka project 2015

Students will select a machine learning project of their choice and use one or more algorithms to solve the selected problem. A good project is such if 1)It is possible to obtain a dataset (either because available or because it can be built) and sufficient information concerning the problem 2)The representation of domain instances (objects relevant for the problem being considered) is not trivial (= sufficiently large number of features, sufficiently large number of values or real-valued features) 3) There is a performance evaluation and 4) (Not strictly necessary but suggested): To solve the problem , students need to learn something new and interdisciplinary

How a project is evaluated:

  • Simple problem, easy-to-model easy-to-describe instances, small dataset, standard ML learning algorithms using WEKA: up to 25/30
  • Simple problem, easy-to-model easy-to-describe instances, medium-large datset, some complex algorithm not in weka or a more efficient implementation of existing algorithm (e.g. some ad-hoc software developed): up to 28
  • Original problem, complex dataset, not straightforward use of algorithms or new algorithm or ad-hoc implementation: up to 30 L

Two very good projects: Will grass grubs damage the corp this year?, A Framework for Genetic Algorithms

Google Group

MANDATORY!!

Please subscribe to this group:

<form action="https://groups.google.com/a/di.uniroma1.it/forum/#!forum/machinelearning2017/boxsubscribe"> <input type=hidden name="hl" value="en"> </form>

 
Subscribe to Machine Learning 2017 @di.uniroma1.it

| Email: |

Visit this group

Slides and course materials (those with date=2016* are UPDATED)

Timetable Topic PPT Details
2016* Introduction   pdf
2016* Inductive Classification   pdf
2016* Introduction to WEKA toolkit 1 ppt pdf
2016* Introduction to WEKA toolkit 2 ppt pdf
2016* Decision Trees ppt pdf
2016** Performance Evaluation: error estimates, confidence intervals, one/two-tail test   pdf
2016* Neural Networks and Deep Learning (denoising autoencoders) ppt pdf
2016* Ensemble methods (bagging, boosting) ppt pdf
2016* Support Vector Machines pptx pdf
2016** Probabilistic learning: Maximum Likelyhood Learning, Naive Bayes ppt pdf
2016* Unsupervised learning: Association Rules ppt pdf
2013 Unsupervised learning: Clustering ppt pdf
2013 Unsupervised learning: Evolutionary methods (Prof. Stefano Nolfi) pdf  
2016* Unsupervised Learning: Reinforcement Learning and Q-Learning ppt pdf
2016* Unsupervised Learning: genetic Algorithms ppt pdf

Syllabus

  • Concept Learning and the General-to-Specific Ordering
  • Decision Tree Learning
  • Evaluation methods: experimental and theoretical methods
  • Ensamble methods
  • Artificial Neural Networks and Deep learning with autoencoders
  • Support Vector Machines
  • Clustering
  • Maximum Likelyhood Learning
  • Data Mining: Apriori algorithm and frequent itemset mining
  • Reinforcement learning and Q-Learning
  • Genetic Algorithms
Topic attachments
I Attachment History Action Size Date Who Comment
Unknown file formatcsv ML2016-LearningSet.csv r1 manage 5079.0 K 2016-03-23 - 15:47 PaolaVelardi  
Edit | Attach | Watch | Print version | History: r220 < r219 < r218 < r217 < r216 | Backlinks | Raw View | Raw edit | More topic actions
Topic revision: r220 - 2017-04-28 - PaolaVelardi






 
Questo sito usa cookies, usandolo ne accettate la presenza. (CookiePolicy)
This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2017 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback