arXiv subjects

Rossi Kamal

Publications and source records attributed to Rossi Kamal.

12 recordsLinked to original sources

arXiv2024

Behave-XAI: Deep Explainable Learning of Behavioral Representational Data

According to the latest trend of artificial intelligence, AI-systems needs to clarify regarding general,specific decisions,services provided by it. Only consumer is satisfied, with explanation , for example, why any classification result is the outcome of any given time. This actually motivates us using explainable or human understandable AI for a behavioral mining scenario, where users engagement on digital platform is determined from context, such as emotion, activity, weather, etc. However, the output of AI-system is not always systematically correct, and often systematically correct, but apparently not-perfect and thereby creating confusions, such as, why the decision is given? What is the reason underneath? In this context, we first formulate the behavioral mining problem in deep convolutional neural network architecture. Eventually, we apply a recursive neural network due to the presence of time-series data from users physiological and environmental sensor-readings. Once the model is developed, explanations are presented with the advent of XAI models in front of users. This critical step involves extensive trial with users preference on explanations over conventional AI, judgement of credibility of explanation.

cs.LG↗

arXiv2024

Deep Recurrent Learning Through Long Short Term Memory and TOPSIS

Enterprise resource planning (ERP) software brings resources, data together to keep software-flow within business processes in a company. However, cloud computing's cheap, easy and quick management promise pushes business-owners for a transition from monolithic to a data-center/cloud based ERP. Since cloud-ERP development involves a cyclic process, namely planning, implementing, testing and upgrading, its adoption is realized as a deep recurrent neural network problem. Eventually, a classification algorithm based on long short term memory (LSTM) and TOPSIS is proposed to identify and rank, respectively, adoption features. Our theoretical model is validated over a reference model by articulating key players, services, architecture, functionalities. Qualitative survey is conducted among users by considering technology, innovation and resistance issues, to formulate hypotheses on key adoption factors.

cs.SE↗

arXiv2024

ETHNO-DAANN: Ethnographic Engagement Classification by Deep Adversarial Transfer Learning

Student motivation is a key research agenda due to the necessity of both postcolonial education reform and youth job-market adaptation in ongoing fourth industrial revolution. Post-communism era teachers are prompted to analyze student ethnicity information such as background, origin with the aim of providing better education. With the proliferation of smart-device data, ever-increasing demand for distance learning platforms and various survey results of virtual learning, we are fortunate to have some access to student engagement data. In this research, we are motivated to address the following questions: can we predict student engagement from ethnographic information when we have limited labeled knowledge? If the answer is yes, can we tell which features are most influential in ethnographic engagement learning? In this context, we have proposed a deep neural network based transfer learning algorithm ETHNO-DAANN with adversarial adaptation for ethnographic engagement prediction. We conduct a survey among participants about ethnicity-based student motivation to figure out the most influential feature helpful in final prediction. Thus, our research stands as a general solution for ethnographic motivation parameter estimation in case of limited labeled data.

cs.CY↗

arXiv2024

LAGAN: Deep Semi-Supervised Linguistic-Anthropology Classification with Conditional Generative Adversarial Neural Network

Education is a right of all, however, every individual is different than others. Teachers in post-communism era discover inherent individualism to equally train all towards job market of fourth industrial revolution. We can consider scenario of ethnic minority education in academic practices. Ethnic minority group has grown in their own culture and would prefer to be taught in their native way. We have formulated such linguistic anthropology(how people learn)based engagement as semi-supervised problem. Then, we have developed an conditional deep generative adversarial network algorithm namely LA-GAN to classify linguistic ethnographic features in student engagement. Theoretical justification proves the objective, regularization and loss function of our semi-supervised adversarial model. Survey questions are prepared to reach some form of assumptions about z-generation and ethnic minority group, whose learning style, learning approach and preference are our main area of interest.

cs.CL↗

arXiv2016

Smart Context-aware Rejuvenation of Engagement on Urban Ambient Augmented Things

The concern over global urbanization trend imposes smart-city as enabling information and communication technology (ICT) to improve urban governance. However, the light trance on better living space is stimulated by socioeconomic impact of escalated senior generation. Hence, ambient assisted living (AAL) emerges for the autonomous provisioning of pervasive things or objects from relevant perturbation for advanced scientific instrumentation. Meanwhile, citizens are observed in being transfixed by lively stimuli of monotonous urban events with the advent of virtual reality or augmented things. Thus, due to the involvement of situation-awareness or contextualization, engagement/participation information as a utility promises to improve urban experience. However, it is complex to grapple meaningful concepts due to personalization obstacles, such as citizen psychology, information gap, service-visualization. Moreover, recommended practices deficit adaptation to monochromatic choice, disparate impairments, mobility and annotation-richness in urban space. Hence, rejuvenation of engagement relates to monitoring and quantification of 'service consumption and graceful degradation' of experience. However, paramount challenges are imposed on this stipulation, such as, unobservability, independence and composite relationship of contexts. Therefore, a parametric Bayesian based model is envisioned to address observability and scalability of contexts and its conjugal relationship with engagement. Last but not the least, systematic framework is demonstrated, which pinpoints key goals of context-aware engagement from participants' opinions, usages and feed-backs.

cs.CY↗

arXiv2016

Parametric Bayesian Rejuvenation in Ambient Assisted Living through Software-based Thematic 5G Management

Ameliorating elderly engagement is vital in rejuvenating independent living. However, recommended practices lack realization of personal traits despite socio -economic promise. The recent proliferation of IoT with the advent of smart-objects/things and personalized services pave the way for context-aware service management. Eventually, the major goal of this paper is to develop a context-aware model in predicting engagement of elderly care. Hence, key requirements are identified for elderly engagement, namely (a) discovery of contexts, which are relevant (b) scaling up (over time) of engagement. However, paramount challenges are imposed on this stipulation, such as, un-observability, independence and composite relationship of contexts. Therefore, a Topic-model based model is proposed to address scalability of contexts and its conjugal relationship with engagement. Eventually, systematic framework is demonstrated, which pinpoints key goals of context-aware services by participants' opinions, usage and feed-back.

cs.CY↗

arXiv2016

Non-Parametric Bayesian Rejuvenation of Smart-City Participation through Context-aware Internet-of-Things (IoT) Management

Tweaking citizen participation is vital in promoting Smart City services. However, conventional practices deficit sufficient realization of personal traits despite socio-economic promise. The recent trend of IoT-enabled smart-objects/things and personalized services pave the way for context-aware services. Eventually, the aim of this paper is to develop a context-aware model in predicting participation of smart city service. Hence, major requirements are identified for citizen participation, namely (a) unwrapping of contexts, which are relevant, (b) scaling up (over time) of participation. However, paramount challenges are imposed on this stipulation, such as, un-observability, independence and composite relationship of contexts. Therefore, a Non-parametric Bayesian model is proposed to address scalability of contexts and its relationship with participation.

cs.CY↗

arXiv2015

Connected Big Data Measurement

In this paper, we have summarized how resilient Big Data monetization scheme outperforms state-of-the art schemes by maintaining a balance between CDS size and routing.

cs.NI↗

arXiv2015

Unsupervised Online Bayesian Autonomic Happy Internet-of-Things Management

In Happy IoT, the revenue of service providers synchronizes to the unobservable and dynamic usage-contexts (e.g. emotion, environmental information, etc.) of Smart-device users. Hence, the usage-context-estimation from the unreliable Smart-device sensed data is justified as an unsupervised and non-linear optimization problem. Accordingly, Autonomic Happy IoT Management is aimed at attracting initial user-groups based on the common interests (i.e. recruitment ), then uncovering their latent usage-contexts from unreliable sensed data (i.e. revenue-renewal ) and synchronizing to usage-context dynamics (i.e. stochastic monetization). In this context, we have proposed an unsupervised online Bayesian mechanism, namely Whiz (Greek word, meaning Smart), in which, (a) once latent user-groups are initialized (i.e measurement model ), (b) usage-context is iteratively estimated from the unreliable sensed data (i.e. learning model ), (c) followed by online filtering of Bayesian knowledge about usage-context (i.e. filtering model ). Finally, we have proposed an Expectation Maximization (EM)-based iterative algorithm Whiz, which facilitates Happy IoT by solving (a) recruitment, (b) revenue-renewal and (c) stochastic- monetization problems with (a) measurement, (b) learning, and (c) filtering models, respectively.

cs.NI↗

arXiv2015

Evolvable Autonomic Management

Autonomic management is aimed at adapting to uncertainty. Hence, it is devised as m-connected k-dominating set problem, resembled by dominator and dominate, such that dominators are resilient up to m-1 uncertainty among them and dominate are resilient up to k-1 uncertainty on their way to dominators. Therefore, an evolutionary algorithm GENESIS is proposed, which resolves uncertainty by evolving population of solutions, while considering uncertain constraints as sub-problems, started by initial populations by a greedy algorithm AVIDO. Theoretical analysis first justifies original problem as NP-hard problem. Eventually, the absence of polynomial time approximation scheme necessitates justification of original problem as multiobjective optimization problem. Furthermore, approximation to Pareto front is verified to be decomposed into scalar optimization sub-problems, which lays out the theoretical foundation for decomposition based evolutionary solution. Finally, case-study, feasibility analysis and exemplary implication are presented for evolvable autonomic management in combined cancer treatment with in-vivo sensor networks.

cs.OH↗

arXiv2015

Resilient Big Data Monetization

Resilient Big Data monetization is devised as k-dominance and m-connectivity problems, such that common-interests are connected by k-ways to measurement tools, which are tied within each other in m-ways. Consequently, a greedy approximation algorithm Plutus (i.e resembling Greek god of wealth) is proposed, which isolates measurement tools to ac-quire domination over common-interests, establishes synergy from common-interests to measurement tools and then acquires divergence and sustains it within measurement tools

cs.NI↗

arXiv2015

Autonomic Resilient Internet-of-Things(IoT)Management

In Resilient IoT, the revenue of service provider is resilient to uncertain usage-contexts(e.g. emotion, environmental contexts) of Smart-device users. Hence, Autonomic Resilient IoT Management problem is decomposed into two subproblems, namely m-connectivity and k-dominance, such that m-alternations on revenue making process is resilient to users common interests, which might be depicted through k-1 alternations of usage-contexts. In this context, a greedy approximation scheme Bee is proposed, which resolves aforementioned sub-problems with five consecutive models, namely Maverick, Siren, Pigmy, Arkeo and Augeas, respectively. Theoretical analysis justifies the problem as NP-hard, combinatorial optimization problem, which is amenable to greedy approximation. Moreover, Bee lays out the theoretical foundation of Resilient Fact-finding, followed by theoretical and experimental(i.e synthetic) proof, which show how Bee-resilience resolves acute CDS measurement problem. Accordingly, experiments on real Social rumor dataset extract dominator and dominate to justify how Bee resilience improves CDS measurement. Finally, case-study and prototype development are performed on Android and Web platforms in a Resilient IoT scenario, where service provider recommends personalized services for Smart-device users.

cs.NI↗