2018/10/9 Dr. Pavel Vařacha（チェコ共和国）の講演会（10月16日（火））のお知らせ
Dr. Pavel Vařachaがチェコ共和国から来日し、10月3日（水）～18日（木）まで、山本佳世子研究室に滞在します。
以下の要領で講演会を開催いたしますので、ご関心のある方は、ぜひご参加ください。
講演会終了後には、意見交換会を開催いたします。
Dr. Pavel Vařacha講演会・意見交換会
日時：10月16日（火）
場所：電気通信大学東2号館402室
講演会：15:00～16:30
意見交換会：17:00～19:00
講演会テーマ
Self‐Organizing Migrating Algorithm and Analytic Programming
Lecture is aimed on two less know machine learning algorithms as well as on their ability to successfully learn and optimize artificial neural network.
1) Self‐Organizing Migrating Algorithm
The lecture describes highly effective evolutionary algorithm SOMA – Self‐Organizing Migrating Algorithm and explains how it can be used to solve various scientific and industrial tasks of optimization. Modern approach of parallelization and distribution of the algorithm is introduced. By this approach SOMA can solve very time demanding problems effectively. SOMA fits on broad class of problems, which can be formulated as a function with one input and arbitrary number of outputs, and generally performing better than older evolutionary algorithms.
2) Analytic Programming
Evolutionary symbolic regression brings excellent aid when scientific or industrial optimization problem cannot be described as a function. The lecture aimed on Analytic Programming as modern, variable and highly effective tools of symbolic regression. Analytic Programming is capable of synthetize resulted function while needs only input and output data. By this approach for example optimal robot path, net structure or approximated cost function can be obtained. Whole process is effectively parallelized to deal with high amount of data within reasonable time.
