Week 3: Research Problems

Announcements

  • Lab 2 is due tomorrow 02/01 by 23:59 PM EST.

  • Lab 3 will be available Wednesday 02/02, it is a team assignment.

  • All Meetings this week are In Person: Lectures, labs, and office hours.

  • Class participation, try EdSTEM and Figma, Group 3 of six groups for lecture notes.

Week 3 Topics

  • Research Problems

  • Title, Keywords, Abstract and Algorithms

  • Your Research Problems

  • Your Research Problems for mid-term

  • Your Research Problems for final

Monday

EEGEyeNet_2021

  • Where to locate Research Problems

  • Paper PDF File

  • Introduction

  • Related Work

  • Discussion and Future Work

  • Conclusion

Craik_2019

  • Where to locate Research Problems

  • Paper PDF File

  • Introduction

  • Results

  • Discussion

  • Conclusion

In-class exercises:

  • Explain to your classmates which research problem(s) you are interested in

  • Post it on Figma

  • Keywords

  • Algorithms

Roy_2019

  • Where to locate Research Problems

  • Paper PDF File

  • Introduction

  • Discussion

  • Conclusion

Lotte_2018

  • Where to locate Research Problems

  • Paper PDF File

  • Introduction

  • Past methods and current challenges

  • Discussion and Future Work

  • Conclusion

Wednesday

Lotte_2018, more details

  • 3. Past methods and current challenges

  • 3.1. A brief overview of methods used 10 years ago

  • LDA and SVM, 'the most popular types, for online and real-time'

  • k-Nearest Neighbour (kNN)

  • Classifier combinations (boosting, voting, or stacking), 'the best performing, in online evaluations.''

  • 3.2. Challenges faced by current EEG signal classification methods

  • 4.5. Other new classifiers

  • 4.5.1. Multilabel classifiers

  • 4.5.2. Classifiers that can be trained from little data

  • sLDA, RF, and the RMDM are 'simple classifiers that are easy to use in practice and provide good results in general, including online.'

  • 5. Discussion and guidelines

  • 5.1. Summary and guidelines

  • 5.2. Open research questions and challenges

  • 6. Conclusion

  • Future work related to EEG-based BCI classification

Algorithms

Andrew Ng's Courses: Machine Learning, Deep Learning, AI

Yann LeCun’s Deep Learning Course: Deep Learning, Course Introduciton

Datasets

Type 1, No Datasets

Type 2, Using Existing Datasets

Type 3, Collect your own Dataset

  • EEGEyeNet_2021

  • Why?

  • Timeline

  • Your projects, mid-term and final

Poster examples AAAI 2021

  • Type 1, No Datasets

  • Type 2, Using Existing Datasets

  • Type 3, Collect your own Dataset

  • AAAI 2021 Posters

  • Research Problems

  • Title

  • Abstract

  • Keywords

  • Algorithms

  • Dataset? Type 1, 2, or 3

Poster Titles

  • Responsible Prediction Making of COVID-19 Mortality (Student Abstract)

  • Unsupervised Causal Knowledge Extraction from Text using Natural Language Inference (Student Abstract)

  • Early Prediction of Children’s Task Completion in a Tablet Tutor using Visual Features (Student Abstract)

  • Fair Stable Matchings Under Correlated Preferences (Student Abstract)

  • Multi-modal User Intent Classification Under the Scenario of Smart Factory (Student Abstract)

  • Incorporating Curiosity into Personalized Ranking for Collaborative Filtering (Student Abstract)

  • An Unfair Affinity Toward Fairness: Characterizing 70 Years of Social Biases in BHollywood (Student Abstract)

  • Is Active Learning Always Beneficial? (Student Abstract)

  • Detection of Digital Manipulation in Facial Images (Student Abstract)

  • Are Chess Discussions Racist? An Adversarial Hate Speech Data Set (Student Abstract)

  • WildfireNet: Predicting Wildfire Profiles (Student Abstract)

In-class exercises:

  • Keywords

  • Algorithms

  • Dataset? Type 1, 2, or 3

Friday

Poster examples AAAI 2020

  • Type 1, No Datasets

  • Type 2, Using Existing Datasets

  • Type 3, Collect your own Dataset

  • AAAI 2020 Posters

  • Research Problems

  • Title

  • Abstract

  • Keywords

  • Algorithms

  • Dataset? Type 1, 2, or 3

Poster Titles

  • Towards an Integrative Educational Recommender for Lifelong Learners (Student Abstract)

  • ESAS: Towards Practical and Explainable Short Answer Scoring (Student Abstract)

  • Leveraging BERT with Mixup for Sentence Classification (Student Abstract)

  • Learning to Classify the Wrong Answers for Multiple Choice Question Answering (Student Abstract)

  • BattleNet: Capturing Advantageous Battlefield in RTS Games (Student Abstract)

  • Travel Time Prediction on Un-Monitored Roads: A Spatial Factorization Machine Based Approach (Student Abstract)

  • MUSIC COLLAB: An IoT and ML Based Solution for Remote Music Collaboration (Student Abstract)

  • Personalized Prediction of Trust Links in Social Networks (Student Abstract)

  • KnowBias: Detecting Political Polarity in Long Text Content (Student Abstract)

vBCI Best Presentation Awards

Oral Presentation Awards

Non-Invasive Category:

  • Biased feedback influences learning in motor imagery BCI training

  • page #11 of 171

Signal Analysis Category

  • Functional connectivity predicts MI-based BCI learning

  • page #6 of 171

Poster Presentation Awards

Non-Invasive Category:

  • Brain-Computer Interfaces for optimal human-machine collaboration

  • page #130 of 171

Signal Analysis Category

  • FReliable outlier detection by spectral clustering on Riemannian manifold of EEG covariance matrix

  • page #51 of 171

Ethical issues

Your Research Problems

  • Title

  • Abstract

  • Keywords

  • Algorithms

  • Dataset? Type 1, 2, or 3