Explainable AI Enabled Infant Mortality Prediction Based on Neonatal Sepsis

Neonatal sepsis is the third most common cause of neonatal mortality and a serious public health problem,especially in developing countries.There have been researches on human sepsis,vaccine response,and immunity.Also,machine learning methodologies were used for predicting infant mortality based on certain features like age,birth weight,gestational weeks,and Appearance,Pulse,Grimace,Activity and Respiration(APGAR)score.Sepsis,which is considered the most determining condition towards infant mortality,has never been considered for mortality prediction.So,we have deployed a deep neural model which is the state of art and performed a comparative analysis of machine learning models to predict the mortality among infants based on the most important features including sepsis.Also,for assessing the prediction reliability of deep neural model which is a black box,Explainable AI models like Dalex and Lime have been deployed.This would help any non-technical personnel like doctors and practitioners to understand and accordingly make decisions.

Aggregator Based RPL for an IoT-Fog Based Power Distribution System with 6LoWPAN

Advanced Metering Infrastructure(AMI)forms an important part in Smart Grids.Routing the data effectively from smart meters to the Edge/Fog node requires an efficient routing protocol.Routing Protocol for Low Power Lossy Area Network(RPL)is a standard routing protocol for IPv6 over Low Power Personal Area Network(6LoWPAN).In a Power Distribution system all the smart meters together form 6LoWPAN network.They communicate with the fog router,which acts as the 6LoWPAN gateway.ContikiRPL was evaluated using Cooja Network simulator for a power distribution network topology.The nodes which were far away from the fog node gave low Packet Delivery Ratio(PDR)and large End to End delay.This paper proposes an aggregation RPL scheme by modifying the existing Contiki RPL.The smart meter nodes communicate to the aggregator,which communicates to the fog node.The results show that the aggregation scheme has 35.6%increase in PDR,lesser hop count and 13.24%decrease in End to End delay on an average compared to existing RPL.

Fuzzy Aggregator Based Energy Aware RPL Routing for IoT Enabled Forest Environment

Forested areas are extremely vulnerable to disasters leading to environmental destruction.Forest Fire is one among them which requires immediate attention.There are lot of works done by authors where Wireless Sensors and IoT have been used for forest fire monitoring.So,towards monitoring the forest fire and managing the energy efficiently in IoT,Energy Efficient Routing Protocol for Low power lossy networks(E-RPL)was developed.There were challenges about the scalability of the network resulting in a large end-to-end delay and less packet delivery which led to the development of Aggregator-based Energy Efficient RPL with Data Compression(CAAERPL).Though CAA-ERPL proved effective in terms of reduced packet delivery,less energy consumption,and increased packet delivery ratio for varying number of nodes,there is still challenge in the selection of aggregator which is based purely on probability percentage of nodes.There has been research work where fuzzy logic been employed for Mobile Ad-hoc Routing,RPL routing and cluster head selection in Wireless Sensor.There has been no work where fuzzy logic is employed for aggregator selection in Energy Efficient RPL.So accordingly,we here have proposed Fuzzy Based Aggregator selection in Energy-efficient RPL for region thereby forming DODAG for communicating to Fog/Edge.We here have developed fuzzy inference rules for selecting the aggregator based on strength which takes residual power,Node degree,and Expected Transmission Count(ETX)as input metrics.The Fuzzy Aggregator Energy Efficient RPL(FA-ERPL)based on fuzzy inference rules were analysed against E-RPL in terms of scalability(First and Half Node die),Energy Consumption,and aggregator node energy deviation.From the analysis,it was found that FA-ERPL performed better than E-RPL.These were simulated using MATLAB and results.