Spring 2017


Fall 2016

For fall 2016, there will be a group meeting every second Thursday at 11:00-12:00 in room A042 (Realfagbygget). The members of the group will give a very quick presentation of their most recent activities and they will present their work schedule for the very next future. The presentation will be done using few slides, possibly following this template: ml_meetings_template.

Additionally, there will be 2 reading sessions each week.

The first reading session is on Tuesday at 12:00-13:00 in room 4.003 (Teknologibygget) and it will be moderated alternatively by Karl Øyvind and Luigi. One week, Karl Øyvind will discuss papers concerning machine learning application on health care. The other week, Luigi will discuss papers on the topic of domain adaptation.

The second reading session is on Thursday at 10:00-11:00 in room A042 (Realfagbygget) and it will be moderated alternatively by Sigurd and Michael. One week, Sigurd will discuss papers concerning kernel methods. The other week, Michael will discuss papers on the topic of deep learning.

Machine Learning in health care (Karl Øyvind)Deep Learning (Michael)Kernel Methods (Sigurd)Domain Adaptation (Luigi)
Exploring patient risk groups with incomplete
knowledge (link)
Auto-Encoding Variational Bayes (link)Mini-Batch Spectral Clustering (link)A New Multivariate Statistical Model for Change Detection in Images Acquired by Homogeneous and Heterogeneous Sensors (link)
Time series representation and similarity based on local autopatterns (link)Multi-Timescale Long Short-Term Memory Neural Network
for Modelling Sentences and Documents (link)
A Dirichlet Process Mixture Model for Spherical Data (link)Graph Matching for Adaptation in Remote Sensing (link)
Learning statistical models of phenotypes using noisy labeled training data (link)Deep Unsupervised Learning through Spatial Contrasting (link)Probabilistic Integration: A Role for Statisticians in Numerical Analysis? (link)Distribution-Matching Embedding for Visual Domain Adaptation (link)
Subjectively Interesting Component Analysis:
Data Projections that Contrast with Prior Expectations (link)
Risk Prediction with Electronic Health Records: A Deep Learning
Approach (link)
Lp-Norm Multiple Kernel Learning (link)Frustratingly Easy Domain Adaptation (link)
Recurrent Neural Networks for Multivariate Time Series with Missing Values (link)Information Theoretic-Learning Auto-Encoder (link)​Multiple Kernel k-Means Clustering with Matrix-Induced Regularization (link)Modelling land cover change in a Mediterranean environment using Random Forests and a multi-layer neural network model (link)

Spring 2016

For spring 2016, Lab seminars are divided into four parts: (1) Paper reading group. (2) Bayesian learning reading group. (3) Deep learning reading group. (4) Machine learning talks.

DatePaper readingDeep learningBayesian learningML talks
Week 23 (06.06)Rogelio Andrade Mancisidor
Industrial PhD: Machine Learning and Santander Bank
Week 21-22 (23.05-30.06)
Traveling, no seminar
Week 20 (16.05)Deep Networks with Stochastic Depth
Gao Huang, Yu Sun, Zhuang Liu, Daniel Sedra, Kilian Weinberger
Week 19 (09.05)Nolinear time-series analysis revisited
Elisabeth Bradley, Holger Kantz
Week 18 (02.05)Lazy Quantum clustering induced radial basis function networks
(LQC-RBFN) with effective centers selection and radii determination
Yiqian Cui, Junyou Shi, Zili Wang
Week 17 (25.04)Deeply supervised nets
Chen-Yu Lee, Saining Xie, Patrick Gallagher, Zhengyou Zhang, Zhuowen Tu
Week 16 (18.04)Spatial Transformer Networks
Max Jaderberg, Karen Simonyan, Andrew Zisserman, Koray Kavukcuoglu
Week 15 (11.04)Tobias Foslid and Filippo Bianchi on the Chinese restaurant process
Week 14 (04.04)Tobias Foslid and Filippo Bianchi on conjugate priors
Week 12 (21.03)P. Orbanz, Lecture Notes on Bayesian Nonparametrics
Week 11 (14.03)Tobias Foslid and Filippo Bianchi on Dirichlet processes
Week 10 (07.03)P. Orbanz, Lecture Notes on Bayesian Nonparametrics
Week 9 (29.02)Filippo Bianchi, Machine Learning @ UiT Lab
Clustering on unstructured data
Week 8 (22.02)S. Gershman and D. Blei. A tutorial on Bayesian nonparametric models. Journal of Mathematical Psychology, 56:1–12, 2012
Week 7 (15.02)M. N. Schmidt, M. Mørup, Non-parametric Bayesian modeling of complex networks, IEEE Signal Processing Magazine,
vol 30(3), pp. 110-128, 2013
Kajsa Møllersen, UiT
Ensemble learning methods for classification
Week 6 (08.02)ADAM: A Method for Stochastic Optimization
Diederik Kingma, Jimmy Lei Ba
Training R-CNNs of various velocities: Slow, fast, and faster
Ross Girshick
Week 5 (01.02)Håvard Vågstøl
Deep Vision
Week 4 (25.01)Fully Convolutional Networks for Semantic Segmentation
Jonathan Long, Evan Shelhamer, Trevor Darrell
Video lecture: Bayesian Nonparametrics 1 - Yee Whye Teh - MLSS 2013 Tübingen
Week 3 (18.01)Deep Manifold Traversal: Changing Labels with Convolutional Features
Jacob R. Gardner, Matt J. Kusner, Yixuan Li, Paul Upchurch, Kilian Q. Weinberger, John E. Hopcroft
Week 2 (11.01)Space-Time Local Embeddings
Ke Sun, Jun Wang, Alexandros Kalousis, Stephane Marchand-Maillet
Deep Residual Learning for Image Recognition
Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun
Week 1 (04.01)Parallel Correlation Clustering on Big Graphs
Xinghao Pan, Dimitris Papailiopoulos, Samet Oymak, Benjamin Recht, Kannan Ramchandran, Michael I. Jordan

Fall 2015

  • 4. Dec. Kajsa Møllersen from the Department of Mathematics and Statistics and the Norwegian Center on Integrated Care and Telemedicine gives a talk: “Choosing the right clustering technique – properties for density-based clustering”.
  • 27. Nov. Associate Professor Matthias Mittner from the Psychology department gives a talk on “When the Brain Takes a Break: A Model-Based Analysis of Mind Wandering”.
  • 13. Nov. Vidar Vikjord and Jan Ove Karlberg gives a presentation about machine learning at Microsoft Development Center Norway.
  • 30. Oct. Karl Øyvind from Machine learning @ UiT talks about his recent work on further developing the so-called Anchor Method [Halpern, Sonntag et al., 2014] for health data analytics for predicting postoperative delirium, a serious condition, after colorectal cancer surgery.
  • 23. Oct. Einar from the Biological Data Processing Lab, Computer Science Department, gives a review of his work on interpreting gene expression data from the NOWAC (kvinner og kreft) cancer study performed by Eiliv Lund’s group at the Med-fak, UiT.
  • 9. Oct. Assoc. Prof. Georg Elvebakk from the Dept. Mathematics and Statistics gives a lecture on Markov chains, Gibbs sampling and the Metropolis-Hastings algorithm.
  • 26. Sept. Raul Primicerio from the Department of Fishery Sciences talks about assessment of vulnerability of ecosystems using machine learning techniques.
  • 11. Sept. Yuan Fuqing from the Department of Engineering and Safety talks about “Building up Machine Learning Model for Condition Monitoring”.
  • 4. Sept: Morten Grønnesby, Biological Data Processing Lab, with the talk “Pulmonary Sound Classification Using Signal Processing and Machine Learning”.
  • 28. August: Machine learning master student Thomas Johansen presents his summer internship at the Norwegian Computing Center, a research institution in Oslo, Norway.

Spring 2015 

  • 24. April:
    ML Presentation: Morten (Master student at Dept. Computer Science) presents the paper “Large Scale Distributed Deep Networks” (; No Reading group this seminar.
  • 17. April:
    ML Presentation: Karl Øyvind introduces random forests and goes in detail on applications such as Microsoft Kinect in the XBox; Reading group: Ch. 7,8 in Criminisi, Shotton and Konukoglu’s Technical Report on Decision Forests (Microsoft).
  • 10. April:
    ML Presentation: Kasper talks about natural language processing and text mining in electronic health records obtained from the University Hospital of North Norway; Reading group: Ch. 6 in Criminisi, Shotton and Konukoglu’s Technical Report on Decision Forests (Microsoft).
  • 27. March:
    ML Presentation: Jonas demonstrates embedding/visualization of songs from Spotify using several different dimensionality reduction methods; Reading group: Ch. 5 in Criminisi, Shotton and Konukoglu’s Technical Report on Decision Forests (Microsoft).
  • 20. March:
    ML Presentation: We watch Nando de Freitas’ lecture on Deep Learning II; Reading group: “Semi-supervised learning with density-based distances,” Bijral et al.
  • 13. March:
    ML Presentation: Robert presents recent work on using composite kernels for fusion of heterogeneous sources in electronic health records, submitted for IEEE JBHI and AMIA; Reading group: Ch. 4 in Criminisi, Shotton and Konukoglu’s Technical Report on Decision Forests (Microsoft).
  • 6. March:
    ML Presentation: Michael presents his Master thesis work on recommender systems; Reading group: Ch. 3 in Criminisi, Shotton and Konukoglu’s Technical Report on Decision Forests (Microsoft).
  • 20. Feb:
    ML Presentation: Katalin presents her ongoing PhD work on Bayesian learning and complexity priors; Reading group: Ch. 3 in Criminisi, Shotton and Konukoglu’s Technical Report on Decision Forests (Microsoft).
  • 13. Feb:
    ML Presentation: We watch Nando de Freitas’ lecture on Deep Learning I; Reading group: Ch. 2 in Criminisi, Shotton and Konukoglu’s Technical Report on Decision Forests (Microsoft).
  • 6. Feb:
    Presentation: Jonas presents a summary of the Big Data Winter School in Tarragona, Spain, with particular focus on Decision trees and Suyken’s kernel methods lecture.

Fall 2014 

  • 25. Nov: Seminar with the UiT High Performance Cluster representative on opportunities for parallel processing and fast computations. Last seminar before Christmas.
  • 18. Nov: Informal joint seminar including Machine Learning @ UiT and the Biological Data Processing Systems Lab.
  • 11. Nov: Thomas presents Cuda and openCL and we discuss GPU processing.
  • 4. Nov: The group discuss Deep Learning, and study some papers from deep
  • 31. Oct: Robert informs about his trip to the UTT (Troyes, France) where he was opponent on a PhD defense. The topic of the defense was the use of Jenssen’s kernel entropy component analysis for flight data monitoring in commercial air traffic, and Robert outlines the main aspect of this very interesting work by Nicolas Chrysanthos.
  • 10. Oct: Cristina presents her latest results on detecting wound infections in electronic health records. Cristina soon leaves for Spain, but is welcome back to Machine Learning @ UiT any time!
  • 3. Oct: Sigurd Løkse presents his results so far on “consensus ranking” using the probabilistic cluster kernel. Very interesting!
  • 26. Sept: Robert talks about the information theoretic approach to machine learning that has been pursued at UiT, and use V. Vikjord, R. Jenssen, “Information Theoretic Clustering using a k-Nearest Neighbor Approach,” vol. 47, Pattern Recognition, 2014, as an example.
  • 19. Sept:  Jonas informs about ideas for code sharing, and how that works. Jonas furthermore  spends a few minutes taking us through his poster for the IEEE MLSP conference in Reims (Jonas and Robert will go there, starts Sunday). We close off the discussion on the topological data analysis paper that we talked about one week ago.
  • 12. Sept: Cristina presents the data set on wound infections and preliminary analysis results. We discuss the paper Lum et al., “Extracting Insights from the Shape of Complex Data using Topology,” Scientific Reports, 2013. Karl Øyvind gives a mini-lecture on Hilbert spaces and Banach spaces.
  • 5. Sept: Karl Øyvind presents his master thesis “Deforming the Vacuum” from the Department of Mathematics and Statistics at UiT. Sigurd explains the idea behind E. Izquierdo-Verdiguier, R. Jenssen, L. Gomez-Chova and G. Camps-Valls, “Spectral Clustering with the Probabilistic Cluster Kernel,” accepted for Neurocomputing, 2014. Sigurd will extend Emma’s work in his master thesis.

Comments are closed.