Student Projects propositions for 2014 Autumn
If you are interested in one of the below projects, Semester Projects or Master Projects, please contact the first person of reference indicated in each description either by telephone, or by email, or by visiting us directly to the LASA offices. If you are looking for a project for 2015 Spring, please click here.
Learning robot optimal trajectories online using Inverse Reinforcement LearningInverse Reinforcement Learning (IRL) is a field of learning the desired objectives, values or rewards by observing the behavior of a system. In this particular scenario, the goal is to learn online the optimal motion parameters of a robot arm, in order to achieve a successful obstacle avoidance. In this project, the user directs the robot motion with a joystick, while the brain activity is monitored. It is shown that an erroneous or unexpected behavior during a task could result in the expression of errorrelated potentials (ErrP) in the brain activity. The aim of this project is the use of ErrPs and/or the feedback from the joystick for predicting online the optimal robot trajectories for obstacle avoidance. The student will study the problem of online prediction of whether a trajectory generated by a given set of parameters will be acceptable to the user, aiming to find a robust classifier that achieves this and converges quickly with a small number of samples. S/he will implement a realtime IRL approach to a Kuka LWR robot arm and gain inhand experience in machine learning and robotic control


Robot teleoperation Combining muscular activity with gazeAn important part of neuroprosthetic control is to decode user’s motion intention. This intention is then converted into appropriate movements for the prosthetic or assistive devices. When controlling prosthetic handarm systems, one can use eye movements as a natural way to determine the object the user intends to grasp. Eye movements give only the direction in which the object of interest may be located but not the exact location. In this project we will examine potentials improvements in localization of the object by fusing gaze detection with monitoring of muscular activity (EMG) of the arm. An estimation of the target position in 2D space would come from the gaze, while EMG could be used to train two machine learning algorithms for regressions to predict the hand position in the x and ydirections. A combination of these two systems is not trivial due to noise introduced by random eye movements, head motion and the nonstationary nature of the EMG signals. The student will gain experience in stateofart computer vision methods as well as machine learning regression methods applied on noisy biomedical signals. The goal of the project is a teleoperation system using machine learning methods, where a user would control remotely a robotic arm and hand.


Robot teleoperation Combining muscular activity with gazeAn important part of neuroprosthetic control is to decode user’s motion intention. This intention is then converted into appropriate movements for the prosthetic or assistive devices. When controlling prosthetic handarm systems, one can use eye movements as a natural way to determine the object the user intends to grasp. Eye movements give only the direction in which the object of interest may be located but not the exact location. In this project we will examine potentials improvements in localization of the object by fusing gaze detection with monitoring of muscular activity (EMG) of the arm. An estimation of the target position in 2D space would come from the gaze, while EMG could be used to train two machine learning algorithms for regressions to predict the hand position in the x and ydirections. A combination of these two systems is not trivial due to noise introduced by random eye movements, head motion and the nonstationary nature of the EMG signals. The student will gain experience in stateofart computer vision methods as well as machine learning regression methods applied on noisy biomedical signals. The goal of the project is a teleoperation system using machine learning methods, where a user would control remotely a robotic arm and hand.


Learning robot optimal trajectories online using Inverse Reinforcement LearningInverse Reinforcement Learning (IRL) is a field of learning the desired objectives, values or rewards by observing the behavior of a system. In this particular scenario, the goal is to learn online the optimal motion parameters of a robot arm, in order to achieve a successful obstacle avoidance. In this project, the user directs the robot motion with a joystick, while the brain activity is monitored. It is shown that an erroneous or unexpected behavior during a task could result in the expression of errorrelated potentials (ErrP) in the brain activity. The aim of this project is the use of ErrPs and/or the feedback from the joystick for predicting online the optimal robot trajectories for obstacle avoidance. The student will study the problem of online prediction of whether a trajectory generated by a given set of parameters will be acceptable to the user, aiming to find a robust classifier that achieves this and converges quickly with a small number of samples. S/he will implement a realtime IRL approach to a Kuka LWR robot arm and gain inhand experience in machine learning and robotic control.


Sparse Solutions for LargeScale Regression Problems
The curse of dimensionality is one of the main challenges in 'Big Data' problems. Unless the learning algorithm has an explicitly imposed sparsity constraint, model complexity will undoubtedly increase with respect to the number of samples. Typical sparse solutions for regression focus on problems where the number of samples "M" is less than the input "P", i.e. "M P", specifically, datasets with > 100,000 samples and where a sparse solution is needed for efficient prediction. Two kernelbased methods exist that are formulated to tackle such problems: 1) Relevance Vector Machines, a Bayesian formulation of Support Vector Machines that applies the Bayesian ‘Automatic Relevance Determination’ (ARD) methodology to linear kernel models. 2) Sparse Gaussian Process with PseudoInputs, whose covariance is parameterized by the locations of "M" pseudoinput points, which we learn by a gradient based optimization (analogous to 'relevant vectors'). Nevertheless, both of these algorithms do not scale to data >100k training samples due to their optimization during training. Based on the literature of 1) and 2), the student should extend one of these algorithms to be capable of handling larger datasets, either by a) reformulating the optimization problem, such that it becomes feasible or b) tackle it with a divideandconquer approach and partition the large dataset into smaller subsets where 1) or 2) can be learned and merging/appropriate aggregation schemes must be introduced. The proposed approach will then be validated on interesting realworld dataset with M > 100k. The solution shall be implemented in Matlab/Python/C++ (the students choice).
[1] Michael Tipping. Relevance vector machine, October 14 2003. US Patent
6,633,857


Towards Incremental Learning: Merging SVMs from independent sample sets
With the increase in data available online and everchanging applications, incremental and online machine learning algorithms that can adapt, learn and unlearn will become essential in the near future. Support Vector Machines (SVM) are undoubtedly one of the most powerful machine learning algorithms to date, however, due to the nature of the posed optimization problem (batch learning), they fall short when applied to incremental/adaptive problems. In this work, we are interested in finding a suitable solution for the problem of "incomplete datasets" or "complementary datasets" for a classification problem. Assume we are given a dataset at a specific point in time and we must learn a model to start predicting immediately. Then, we are suddenly given a new set of samples which belong to the same dataset. The question now is: What do we do with this new data? Do we relearn the entire model with all the datapoints? What if the samples are contradictory? Can we learn a new decision function from the new samples and merge them to the old model, without hindering performance on classification? Can we incrementally update the old model with our new samples? What if we suddenly realize that some samples were labeled erroneously and we would like to 'unlearn' them? These are the [subset of] questions that the student should try to answer. Seldom work in SVM literature is capable of handling these issues. The few works that can, are categorized into 1) "active/online methods" where training points are fed onebyone and the SVM is learned sequentially [1] and 2) "ensemble methods" where a dataset is 'partitioned' into Nsets where NSVMs are learned and basic aggregation schemes are applied to generate a final machine [2]. These approaches, however, are mostly suitable for handling large datasets and focus primarily on improving training time (i.e. efficient learning). By leveraging ideas from 1), 2) and online convex optimization [3], the student must propose an efficient and adaptable SVM learning scheme capable of solving all [or a subset] of the issues imposed by the proposed incremental learning problem. The solution shall be implemented in Matlab/Python/C++ (the students choice).
[1] Antoine Bordes, Seyda Ertekin, Jason Weston and Léon Bottou: Fast Kernel Classifiers with Online and Active Learning, Journal of Machine Learning Research, 6:15791619, September 2005.

