Personal robotic assistants help reducing the manual efforts being put by humans in their day-to-day tasks. In this paper, we develop a voice-controlled personal assistant robot. The human voice commands are given to the robotic assistant remotely, by using a smart mobile phone. The robot can perform different movements, turns, start/stop operations and relocate an object from one place to another. The voice commands are processed in real-time, using an online cloud server. The speech signal commands converted to text form are communicated to the robot over a Bluetooth network. The personal assistant robot is developed on a micro-controller based platform and can be aware of its current location. The effectiveness of the voice control communicated over a distance is measured through several experiments. Performance evaluation is carried out with encouraging results of the initial experiments. Possible improvements are also discussed towards potential applications in home, hospitals and industries.Index Terms-smart assistant robot, control over voice, Bluetooth, Android based smart devices.
Due to potentially large number of applications of real-time data stream mining in scientific and business analysis, the real-time data streams mining has drawn attention of many researchers who are working in the area of machine learning and data mining. In many cases, for real-time data stream mining online learning is used. Environments that require online learning are non-stationary and whose underlying distributions may change over time i.e. concept drift, because of which mining of real-time data streams with concept drifts is quite challenging. However, ensemble methods have been suggested for this particular situation. This paper reviews various online methods of drift detection. We also present some results of our experiments that show the comparison of some online drift detection (concept drift) methods.
In the real wo rld, most of the applications are inherently dynamic in nature i.e. their underlying data distribution changes with time. As a result, the concept drifts occur very frequently in the data stream. Concept drifts in data stream increase the challenges in learning as well, it also significantly decreases the accuracy of the classifier. However, recently many algorith ms have been proposed that exclusively designed for data stream mining while considering drift ing concept in the data stream.This paper presents an empirical evaluation of these algorith ms on datasets having four possible types of concept drifts namely; sudden, gradual, incremental, and recurring drifts.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.