Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

https://www.healio.com/psychiatry/journals/psycann/1976-12-6-12/{806b8887-c02e-40c8-a228-57f1cdb2bdba}/the-stress-of-life.pdf

The stress of life

The corporation of the 1990s: Information technology and organizational transformation

/pdf/becoming-critical-education-knowledge-and-action-research-v9w7t6i8zs.pdf

Becoming Critical: education, knowledge and action research

It represents a universally applicable attitude and skill set everyone, not just computer scientists, would be eager to learn and use.

/pdf/computational-thinking-ji-suan-lun-de-si-kao-3qvv34ha3s.pdf

Computational Thinking 計算論的思考

The aim of this paper is double: (a) to record the latest theoretical considerations (literature review) in the field of STEM (acronym of Science, Technology, Engineering, Mathematics), Educational Robotics and the Educational Robotic Platforms used in their implementation, and (b) to validate the argumentation on the potential contribution of an Action Research implementation on STEM education with the ultimate goal of designing and developing an “open philosophy”, low-cost, hardware and software educational platform for the implementation of STEM and Educational Robotics. This paper is divided into 7 sections: Introduction, STEM Education, Educational Robotics, Problem statement, Action Research, Methodology, and Conclusion. The Introduction introduces the concept and necessity of STEM education approach. STEM Education section reviews recently published scientific literature related to STEM education (literature review) and summarize the pros and barriers of its use in education. Educational Robotics introduces the robotics as an educational tool and presents empirical evidence on its effectiveness. Educational Robot Platforms subsection presents the most popular -along with their main specseducational robots for STEM and Educational Robotics use. Problem statement section identifies the scientific gap and composes the necessity to implement research (specifically an Action Research) on designing and developing an “open philosophy”, low-cost, hardware and software academic platform for the implementation of STEM and Educational Robotics. Action research section reviews recently published scientific literature related to action research. Research Methodology section presents research’s proposal development phases and finally, Conclusion summarizes paper’s findings.

/pdf/action-research-implementation-in-developing-an-open-source-2j3qjd820n.pdf

Action Research Implementation in Developing an Open Source and Low Cost Robotic Platform for STEM Education

One of the numerous implementations that the Intelligent Systems Lab of the Technological Institute of Piraeus, Greece, has recently accomplished is described in this paper. Within the area of hand-on autonomous robotics, the project combines computer science in modern wireless network protocol communication, with microcontroller-based intelligent motion-control tasks. The design of the presented pair of versatile robots is intended for a "predator and prey" application domain. Both robot designs are based on low cost and easy to find equipment, though modern and up-to-date technology. Two different multitasking and flexible algorithms are designed in order to investigate their performance and compare the outcomes. In particular, an intelligent algorithm based on data capturing and processing within a tolerable elapsed time is presented, with two different performance scenarios explained and compared in detail.

/pdf/intelligent-algorithms-based-on-data-processing-for-modular-44lojfo1ly.pdf

Intelligent Algorithms Based on Data Processing for Modular Robotic Vehicles Control

In every business, the production of knowledge, coming from the process of effective information, is recognized as a strategic asset and source of competitive advantage. In the field of railways, a vast amount of data are produced, which is necessary to be assessed, deployed in an optimum way, and used as a mechanism, which will lead to making the right decisions, aiming at saving resources and maintain the fundamental principle of the railways which is the passengers’ safety. This paper uses stored-inactive data from a Greek railway company, and uses the method of data mining and applies machine learning techniques to create strategic decision support and draw up a risk and control plan for trains. We make an effort to apply Machine Learning open source software (Weka) to the obsolete procedures of maintenance of the rolling stock of the company (hand-written work orders from the supervisors to the technicians, dealing with the dysfunctions of a train unit by experience, the lack of planning and coding of the malfunctions and the maintenance schedule). Using the J48 and M5P algorithms from the Weka software, data are recorded, processed, and analyzed that can help monitor or discover, with great accuracy, the prevention of possible damage or stresses, without the addition of new recording devices—monitoring on trains, with the aim of predicting the diagnosis of the train fleet. The innovative method is capable of being used as a tool for the optimization of the management’s performance of the trains to provide the appropriate information for the implementation of planning and the technical ability of the trains in order to achieve the greatest target of importance for the railways, which is the passengers’ safety.

Predictive Maintenance Using Machine Learning and Data Mining: A Pioneer Method Implemented to Greek Railways

The increase of human needs has pushed the robotics sector to evolve even more. One area that begins to bloom in robotic systems are semi-autonomous or autonomous monitoring and guarding vehicles from a distance. The present platform is an example of such a system. It will be the analysis of its design, construction and operation. The operator driving the vehicle will have video feed and additional information about the space monitor. This system could be the basis of a complete automated platform with more sensors that will cover more needs.

/pdf/design-of-an-autonomous-robotic-vehicle-for-area-mapping-and-xoa1mrocpm.pdf

Design of an Autonomous Robotic Vehicle for Area Mapping and Remote Monitoring

Olfaction can enhance the experience of music, films, computer games and virtual reality applications. However, this area is less explored than other areas such as computer graphics and audio. Most advanced olfactory displays are designed for a specific experiment, they are hard to modify and extend, expensive, and/or can deliver a very limited number of scents. Additionally, current-generation olfactory displays make no decisions on if and when a scent should be released. This paper proposes a low-cost, easy to build, powerful smart olfactory display, that can release up to 24 different aromas and allow control of the quantity of the released aroma. The display is capable of absorbing back the aroma, in an attempt to clean the air prior to releasing a new aroma. Additionally, the display includes a smart algorithm that will decide when to release certain aromas. The device controller application includes releasing scents based on a timer, text in English subtitles, or input from external software applications. This allows certain applications (such as games) to decide when to release a scent, making it ideal for gaming. The device also supports native connectivity with games developed using a game development asset, developed as part of this project. The project was evaluated by 15 subjects and it was proved to have high accuracy when the scents were released with 1.5 minutes’ delay from each other.

Michail Papoutsidakis

Papers

Action Research Implementation in Developing an Open Source and Low Cost Robotic Platform for STEM Education

Intelligent Algorithms Based on Data Processing for Modular Robotic Vehicles Control

Predictive Maintenance Using Machine Learning and Data Mining: A Pioneer Method Implemented to Greek Railways

Design of an Autonomous Robotic Vehicle for Area Mapping and Remote Monitoring

Towards Smart Gaming Olfactory Displays.