|
|
Bridging the Gap: Towards Unified Approach to Perfect and Imperfect Information Games
Moravčík, Matej ; Hladík, Milan (advisor) ; Pilarski, Patrick (referee) ; Kroupa, Tomáš (referee)
From the onset of AI research, games have played an important part, serving as a benchmark for progress in artificial intelligence. Recent approaches using search in combination with learning from self-play have shown strong performance and the ability to generalize across a wide range of perfect information games. In contrast, the leading algorithms for imperfect information traditionally used a small, abstract version of a game and solved this abstraction in one go. This thesis introduces a chain of improvements for imperfect information algorithms that culminates in two significant milestones that helped bridge the gap between perfect and imperfect information games. The first milestone is DeepStack - the first agent that successfully used a combination of sound search and a learned value function in imperfect information games. This led to the first AI to achieve victory over human professional players in no-limit poker. The second milestone is Player of Games - a universal algorithm that can master both perfect and imperfect information games starting from scratch. 1
|
|
| |
|
|
Production of plough share screw
Hladík, Milan ; Jopek, Miroslav (referee) ; Forejt, Milan (advisor)
The bachelor's thesis presents a design of plow screw production technology used to attach the blade to the plow frame. The screw is made of structural steel 14,140, in a production series of 120,000 pcs / year. From the possible technologies, the technology of volume cold forming was chosen. From the literature sources of the issue of volume cold forming, the optimal procedure of screw production was proposed. With the help of design, technological and control calculations, a machine was designed and subsequently a production tool, for which drawing documentation was prepared. With the help of technical - economic evaluation, the costs of component production were compared using volume forming and machining.
|
|
| |
|
|
Multiobjective shortest path problem with interval costs
Březina, Jiří ; Hladík, Milan (advisor) ; Fink, Jiří (referee)
The multiobjective shortest path problem with interval costs is a gener- alization of the single-pair shortest path problem. In this problem, the edge weights are represented as tuples of intervals. The aim is to find the path that minimizes the maximum regret. We present theorems regarding the compu- tation of the regret and the efficiency of a feasible solution to the problem. The main result of the thesis is an algorithm seeking for the solution with the least regret in the interval multiobjective shortest path problem. 1
|
|
|
Geometry of interval linear systems
Kotecký, Cyril ; Hladík, Milan (advisor) ; Rada, Miroslav (referee)
Interval linear systems of equalities and inequalities are linear systems, where the real numbered entries of the vectors and matrices are replaced with closed intervals of real numbers. The sets of solutions to these systems have interesting properties, mainly that they are unions of exponentially many convex polyhedra. This makes solving many problems hard, while on the other hand, the solution sets have a form that is convenient to analyze. This thesis deals with studying the geometry of such sets. We will begin by reviewing known properties of these sets, such as boundedness and connectedness. But mostly, we will focus on the conditions for convexity and the characterization of the convex hull, which are both known for the special case of systems with invertible interval matrices. Using polyhedral theory, we will broaden these results, mostly to general systems of interval linear inequalities. We will present illustrative examples, some serving as counter-examples in cases where generalizations are not possible. 1
|
|
|
Determination of risks, benefits and efficiency in the implementation of the information manufacturing system MES in production
Hladík, Milan ; Luňáček, Jiří (referee) ; Osička, Karel (advisor)
This master thesis deals with the analysis of the possibilities of implementing the MES manufacturing information system in a company that focuses on the production of medical devices. The thesis is divided into several parts. Firstly, the theoretical background of the thesis is described and then the current situation in the company is mapped out in detail, especially the area of using paper documentation, the so-called DHR. In the practical part of the thesis, the time and financial savings for the company in case of implementation of the system are then quantified, as well as the cost of the investment. Subsequently, the benefits and risks that may arise in such a large-scale project are identified, and at the same time measures are proposed to eliminate these risks completely or reduce their impact. It also includes a proposal for the time allocation necessary for implementation, including the division into individual phases in which the system would be deployed for production. The thesis concludes with a calculation of the return on investment and a proposal for a solution.
|
|
| |
|
|
Results of the autumn extended research of the ichthyofauna of the lakes Laka and Čertovo, including the adjacent streams.
Blabolil, Petr ; Bartoň, Daniel ; Draštík, Vladislav ; Jůza, Tomáš ; Hladík, Milan ; Holubová, Michaela ; Hůda, J. ; Kočvara, Luboš ; Muška, Milan ; Peterka, Jiří ; Vrba, Jaroslav
The report describes the results of a unique survey of the ichthyofauna in two areas of the Šumava National Park, where catches were made by an electric generator in the autumn of 2021.
|
|
|
Search in Imperfect Information Games
Schmid, Martin ; Hladík, Milan (advisor) ; Szepesvari, Csaba (referee) ; Bošanský, Branislav (referee)
From the very dawn of the field, search with value functions was a fun- damental concept of computer games research. Turing's chess algorithm from 1950 was able to think two moves ahead, and Shannon's work on chess from 1950 includes an extensive section on evaluation functions to be used within a search. Samuel's checkers program from 1959 already combines search and value functions that are learned through self-play and bootstrapping. TD-Gammon improves upon those ideas and uses neural networks to learn those complex value functions - only to be again used within search. The combination of decision-time search and value functions has been present in the remarkable milestones where computers bested their human counterparts in long standing challenging games - DeepBlue for Chess and AlphaGo for Go. Until recently, this powerful framework of search aided with (learned) value functions has been limited to perfect information games. As many interesting problems do not provide the agent perfect information of the environment, this was an unfortunate limitation. This thesis introduces the reader to sound search for imperfect information games. 1
|
guest ::
