My research project
Time-series Representation and Analysis Using Machine Learning Techniques
I am currently starting the second year of my PhD on "Time-series Representation and Analysis Using Machine Learning Techniques" at the Centre for Vision, Speech and Signal Processing (CVSSP), University of Surrey. My supervisor is Professor Payam Barnaghi and my co-supervisor is Dr Shirin Enshaeifar.
My research is about representing time-series data in patterns and analysing these patterns using Machine Learning techniques. The aim is to gain insights and knowledge from time-series data which can be used in different domains; health-care, finance, multimedia, smart cities and etc. With this information, we are going to understand the previous trend of data in addition to predicting the future trend of it which can be used in strategic planning.
University roles and responsibilities
- PhD Candidate on "Time-series Representation and Analysis Using Machine Learning Techniques"
- Task Leader for EU-H2020 IoTCrawler Project with focus on developing a search engine for Internet of Things devices, funded by the European Commission
With the proliferation of sensors and IoT technologies, stream data are increasingly stored and analysed, but rarely combined, due to the heterogeneity of sources and technologies. Semantics are increasingly used to share sensory data, but not so much for annotating stream data. Semantic models for stream annotation are scarce, as generally, semantics are heavy to process and not ideal for Internet of Things (IoT) environments, where the data are frequently updated. We present a light model to semantically annotate streams, IoT-Stream. It takes advantage of common knowledge sharing of the semantics, but keeping the inferences and queries simple. Furthermore, we present a system architecture to demonstrate the adoption the semantic model, and provide examples of instantiation of the system for different use cases. The system architecture is based on commonly used architectures in the field of IoT, such as web services, microservices and middleware. Our system approach includes the semantic annotations that take place in the pipeline of IoT services and sensory data analytics. It includes modules needed to annotate, consume, and query data annotated with IoT-Stream. In addition to this, we present tools that could be used in conjunction to the IoT-Stream model and facilitate the use of semantics in IoT.
The rapid growth of Internet of Things (IoT) and sensing technologies has led to an increasing interest in time-series data analysis. In many domains, detecting patterns of IoT data and interpreting these patterns are challenging issues. There are several methods in time-series analysis that deal with issues such as volume and velocity of IoT data streams. However, analysing the content of the data streams and extracting insights from dynamic IoT data is still a challenging task. In this paper, we propose a pattern representation method which represents time-series frames as vectors by first applying Piecewise Aggregate Approximation (PAA) and then applying Lagrangian Multipliers. This method allows representing continuous data as a series of patterns that can be used and processed by various higher-level methods. We introduce a new change point detection method which uses the constructed patterns in its analysis. We evaluate and compare our representation method with Blocks of Eigenvalues Algorithm (BEATS) and Symbolic Aggregate approXimation (SAX) methods to cluster various datasets. We have also evaluated our proposed change detection method. We have evaluated our algorithm using UCR time-series datasets and also a healthcare dataset. The evaluation results show significant improvements in analysing time-series data in our proposed method.
Due to the rapid development of the Internet of Things (IoT) and consequently, the availability of more and more IoT data sources, mechanisms for searching and integrating IoT data sources become essential to leverage all relevant data for improving processes and services. This paper presents the IoT search framework IoTCrawler. The IoTCrawler framework is not only another IoT framework, it is a system of systems which connects existing solutions to offer interoperability and to overcome data fragmentation. In addition to its domain-independent design, IoTCrawler features a layered approach, offering solutions for crawling, indexing and searching IoT data sources, while ensuring privacy and security, adaptivity and reliability. The concept is proven by addressing a list of requirements defined for searching the IoT and an extensive evaluation. In addition, real world use cases showcase the applicability of the framework and provide examples of how it can be instantiated for new scenarios.
The rapid growth in collecting and sharing sensory observation form the urban environments provides opportunities to plan and manage the services in the cities better and allows citizens to also observe and understand the changes in their surrounding in a better way. The new urban sensory data also creates opportunities for further application and service development by creative industries and start-ups. However, as the size and diversity of this data increase, finding and accessing the right set of data in a timely manner is becoming more challenging. This paper describes a search engine designed for indexing, searching and accessing urban sensory data. We present the key feature and architecture of the system and demonstrate some of the functionalities that are provided by searching for raw sensory observations and also pattern search functions that are enabled by a pattern analysis algorithm, supported by monitoring of data streams for changes in quality of information and remediation.
Edge computing can improve the scalability and efficiency of IoT systems by performing some of the analysis and operations on the nodes or on intermediary edge devices. This will reduce the energy consumption, data transmission load and latency by shifting some of the processes to the edge devices. In this paper, we introduce a pattern extraction method which uses both the Lagrangian Multiplier and the Principal Component Analysis (PCA) to create patterns from raw sensory data. We have evaluated our method by applying a clustering method on constructed patterns. The results show that by using our proposed Lagrangian-based pattern extraction method, the existing clustering algorithms perform more accurately - by up to 20% higher compared with the state-of-the-art methods, especially in dealing with dynamic real-world data. We have conducted our evaluations based on synthetic and real-world data sets and have compared the results to the existing state-of-the-art approaches. We also discuss how the proposed methods can be embedded into the edge computing devices in IoT systems and applications.
In recent years, the development and deployment of Internet of Things (IoT) devices has led to the generation of large volumes of real world data. Analytical models can be used to extract meaningful insights from this data. However, most of IoT data is not fully utilised, which is mainly due to interoperability issues and the difficulties to analyse data collected by heterogeneous resources. To overcome this heterogeneity, semantic technologies are used to create common models to share various data originated from heterogeneous sources. However, semantics add further overhead to data delivery, and the processing time to annotate the data with the model can increase the latency and complexity in publishing and querying the annotated data. In this paper, we present a lightweight semantic model to annotate IoT streams. The metadata descriptions that are provided in the models are used for search and discovery of the data using various attributes such as value and type. The proposed model extends commonly used ontologies such as W3C/OGC SSN ontology and its recent lightweight core, SOSA, and includes concepts to describe streaming IoT data. We also show use cases, tools and applications where the proposed model has been used.