Search

Personalized federated learning (PFL) aims to train model(s) that can perform well on the individual edge-devices' data where the edge-devices (clients) are usually IoT devices like our mobile phones. The participating clients for cross-device settings, in general, have heterogeneous system capabilities and limited communication bandwidth. Such practical properties of the edge-devices, however, are overlooked by many recent work in PFL, which use the same model architecture across all clients and incur high communication cost by directly communicating the model parameters. In our work, we propose a novel and practical PFL framework named COMET where clients can use heterogeneous models of their own choice and do not directly communicate their model parameters to other parties. Instead, COMET uses clustered codistillation, where clients use knowledge distillation to transfer their knowledge to other clients with similar data distributions. This presents a practical PFL framework for the edge-devices to train through IoT networks by lifting the heavy communication burden of communicating large models. We theoretically show the convergence and generalization properties of COMET and empirically show that COMET achieves high test accuracy with several orders of magnitude lower communication cost while allowing client model heterogeneity compared to the other state-of-the-art PFL methods.

This paper reviews the literature on the barriers encountered by consumers while purchasing energy efficient household appliances in the last decade. Energy efficient household appliances are the appliances which consume minimum energy while functioning at their fullest capacity. It is observed that the studies on barriers to energy efficient appliances are highly prevalent in the developed economies while there are not many in the BRICS nations. Moreover, in the developed nations, the barriers towards efficient appliances has been analysed from the lens of behavioural economics disregarding the rational economic perspective, while studies of similar kind are lacking in the emerging economies. The review proposes a model based on theory of planned behaviour and Prospect theory to predict the impact of barriers on consumer’s purchase behaviour. it recommends that the linear relationship between purchase intention and purchase behaviour is mediated by the presence of barriers like perceived product risks, subjective norm, perceived personal inconvenience and price sensitivity.

The neoclassical model of economics assumes that greater the information available, the easier it is for a consumer to make decisions. In the context of electricity usage, overwhelming amounts of information presented on energy bills and a lack of real-time information deter consumers from understanding their own energy usage and developing personalized strategies to reduce it. This study assesses the literacy levels regarding electricity bills amongst consumers in India. Regression results indicate that homes with stronger beliefs in energy conservation minimize their usage. The size of the dwelling,demographics, and the ownership of the dwelling influence literacy levels regarding electricity bills. The study had a counterintuitive finding, as the respondents were found reluctant to receive a normative comparison of their electricity bills—strange for a collectivistic cultured society like India. The study recommends making electricity bills more visually appealing and increasing their frequency to increase awareness of energy usage.

Increasing environmental concerns have resulted in consumers indulging in conserving energy by curtailing energy usage rather than switching to energy efficient practices. Energy efficient appliances (EEAs) offer the dual advantages of appliance usage with reduced energy consumption. Research suggests that use of such appliances, in residences is limited. This study was thus undertaken to understand individual resident's awareness towards EEAs and factors influencing their purchase. A qualitative approach consisting of 31 personal interviews were conducted with individuals from socio-economic categories A1, A2 and B1. While awareness of EEAs was present among respondents; benefits of using such appliances were neither widely known nor easily quantifiable. Purchase decisions were ruled by appliance price but did not deter adoption, the availability of appliances in desired aesthetics and family involvement in purchase decision. Higher penetration of such appliances can be ensured if consumers can estimate exact savings in monthly electricity bills due to EEAs at the time of purchase.

The latest Disney channel commercial for the Indian market depicts the move of Disney from being just a kid's channel to a family channel. This move signals the importance of “Families” in the Indian settings. In the Indian context, family has been traditionally considered an important decision-making unit with respect to purchasing products and services required for the household on a daily basis; however, there have not been in depth studies understanding the family as a decision unit and the changes this unit has witnessed since the opening of economy in 1990s. This article compares and contrasts the changing purchase decision making of millennial generation with their parents. Qualitative interviews are used to gather data for family decision making in addition to exclusive secondary sources of data like research papers, articles, and books on the Indian economy and the Indian family structure. Literature from the field of sociology and political science is also referred to, to understand the socio-political and socio-economic impacts on the consumer.

We present here a modified, sonication-assisted transient transformation assay for rapid analysis of cis-regulatory elements. We tested promoter elements from MIR159Blocus of Brassica junceaby generating stable transgenic lines and compared the transcriptional activity of GUS reporter with that of the transient assay method. To obtain reliable and repeatable results, and to omit false-positive data, we optimized several parameters including sonication duration and cycle and concentration of Agrobacterium tumefaciensmeasured as optical density (O.D.) at 600 nm. To the best of our knowledge, this is the first report of promoter characterization of MIR159B from Brassica juncea, and comparative analysis of stable and transient lines. Our analysis shows that the protocol described herein allows understanding promoter activity/transcriptional control in tissues other than leaf or protoplast which have remained the mainstay for transient analysis thus far. We tested reporter gene GUS under the control of constitutive promoter, CaMV 35S, and MIR159bfrom Brassica juncea. We optimized the duration of sonication (5-, 10- and 15-min cycle), bacterial density (measured as O.D at 600 nm = 0.6/0.8/1.0) and Agro-infection time (5, 10, 15 min), and co-cultivation (12-, and 24-h). Sonication cycle of 10-min, followed by Agro-infection and co-cultivation with Agrobacterium tumefacienswith O.D. 600 nm = 0.8 and for 12 h was found to be optimum. We could successfully express reporter genes in deep-seated tissues such as floral organs and pollen grains where it was previously not possible to perform transient assay. Constitutive GUS activity was observed when reporter was placed under control of the constitutive promoter of CaMV 35S. Reporter GUS when placed under transcriptional control of MIR159b promoter from Brassica junceashowed reporter activity in floral tissues, in mature pollen grains. Comparative analysis of reporter activity from stable transgenic lines at T2 generation with that of transient assay system reveals identical to near-identical reporter activity. Transient assay could be successfully performed in tissues collected not only from Arabidopsis thaliana, but also from Brassica junceaand Brassica nigrato demonstrate its wide applicability. Our modified method thus has the potential of quick and rapid analysis of promoter activity and allows us to record the developmental dynamics and spatio-temporal expression pattern driven by specific promoters. Suitable modification and controls should also allow analysis of hormonal regulation and identification of trans-factors via DNA–protein interactions. Furthermore, this method can also be extended to study promoters under various environmental conditions that otherwise do not allow growth and complete life cycle of healthy plants and can be modified to test reporter activity in other non-model plants or plants with long life cycle.

The maximum possible throughput (rate of task completion) of a multi-server system is typically the sum of the service rates of individual servers. Recent works show that task replication can boost the throughput, in particular if the service time has high variability (Cv > 1). Thus, redundancy can be used to create synergy among servers such that their overall throughput is greater than sum of individual servers. This paper seeks to find the fundamental limit of this capacity boost achieved by task replication. The optimal adaptive replication policy can be found using a Markov Decision Process (MDP) framework, but the MDP is hard to solve in general. We propose two replication policies, MaxRate and AdaRep that gradually add replicas only when needed. To quantify the optimality gap of these policies, we also derive an a upper bound on the service capacity for the two-server case.

The purpose of this paper is to determine the extent of environmental concern among consumers while purchasing personal care products. This paper also examines the role of consumer's willingness to pay for environmentally friendly products while making green purchase. It also tests if willingness to pay can be predicted based on certain product attributes. An exploratory factor analysis was carried out to develop the conceptual framework and based on which a questionnaire was designed. Primary data was collected from 300 respondents, women working in Information technology sector between the age group 25-45. The results show that there is high amount of environmental awareness among the consumers. However due to lack of proper and reliable information exchange between the manufacturers and the consumers this awareness does not materialize in green purchase behavior. Consumer's willingness to pay extra for ecolabeled product is dependent on the product reliability and appeal. Hence ecolabeled products stand a chance to be selected if they are appealing and reliable.

In cloud computing jobs consisting of many tasks run in parallel, the tasks on the slowest machines (straggling tasks) become the bottleneck in the completion of the job. One way to combat the variability in machine response time is to add replicas of straggling tasks and wait for the earliest copy to finish. Using the theory of extreme order statistics, we analyze how task replication reduces latency, and its impact on the cost of computing resources. We also propose a heuristic algorithm to search for the best replication strategies when it is difficult to model the empirical behavior of task execution time and use the proposed analysis techniques. Evaluation of the heuristic policies on Google Trace data shows a significant latency reduction compared to the replication strategy used in MapReduce.

Despite the promise of supported eText (or digital text) for the education of students with visual impairments, there is little research on this topic in general and even less in the domain of mathematics. Using a qualitative approach, this study explored the current system used by students with visual impairments to access algebra and the impact of a supported eText player (ReadHear™) on their access to and understanding of algebraic expressions. The findings suggested that students were successful with using the technology to access and understand algebraic expressions. The findings also indicated that efficiency and individualization were key aspects to students' use of and appreciation for the tool, which differs from students' traditional means of accessing mathematics. Additional research is needed to understand how supported eText compares to traditional methods that students with visual impairments use to access text and how this technology can be used in the mathematics classroom.

This project explored a newly developed computer-based voice input, speech output (VISO) calculator. Three high school students with visual impairments educated at a state school for the blind and visually impaired participated in the study. The time they took to complete assessments and the average number of attempts per problem were recorded when they used the VISO calculator and their typical individualized method of calculation. The results indicated that the time to complete assessments and the average number of attempts per problem decreased as the students continued to use the VISO calculator. Students reported positive perceptions of the calculator, particularly noting the independence it provided as compared to their typical means of calculation. Implications for practice, limitations of the study, and future research directions are discussed.

The author relates her mother's concern about her future. The author says her mother does not want her to marry an American boy for they do not know their culture and traditions. She states that her mother wants her to marry an Indian boy, a Maharastrian born and raised in the U.S. so that she will not have to worry about his citizenship.