Traveling Salesman Problems (TSPs) have been a long-lasting interesting challenge to researchers in different areas. The difficulty of such problems scales up further when multiple objectives are considered concurrently. Plenty of work in evolutionary algorithms has been introduced to solve multi-objective TSPs with promising results, and the work in deep learning and reinforcement learning has been surging. This paper introduces a multi-objective deep graph pointer network-based reinforcement learning (MODGRL) algorithm for multi-objective TSPs. The MODGRL improves an earlier multi-objective deep reinforcement learning algorithm, called DRL-MOA, by utilizing a graph pointer network to learn the graphical structures of TSPs. Such improvements allow MODGRL to be trained on a small-scale TSP, but can find optimal solutions for large scale TSPs. NSGA-II, MOEA/D and SPEA2 are selected to compare with MODGRL and DRL-MOA. Hypervolume, spread and coverage over Pareto front (CPF) quality indicators were selected to assess the algorithms’ performance. In terms of the hypervolume indicator that represents the convergence and diversity of Pareto-frontiers, MODGRL outperformed all the competitors on the three well-known benchmark problems. Such findings proved that MODGRL, with the improved graph pointer network, indeed performed better, measured by the hypervolume indicator, than DRL-MOA and the three other evolutionary algorithms. MODGRL and DRL-MOA were comparable in the leading group, measured by the spread indicator. Although MODGRL performed better than DRL-MOA, both of them were just average regarding the evenness and diversity measured by the CPF indicator. Such findings remind that different performance indicators measure Pareto-frontiers from different perspectives. Choosing a well-accepted and suitable performance indicator to one’s experimental design is very critical, and may affect the conclusions. Three evolutionary algorithms were also experimented on with extra iterations, to validate whether extra iterations affected the performance. The results show that NSGA-II and SPEA2 were greatly improved measured by the Spread and CPF indicators. Such findings raise fairness concerns on algorithm comparisons using different fixed stopping criteria for different algorithms, which appeared in the DRL-MOA work and many others. Through these lessons, we concluded that MODGRL indeed performed better than DRL-MOA in terms of hypervolumne, and we also urge researchers on fair experimental designs and comparisons, in order to derive scientifically sound conclusions.

As we delve into a life governed by Artificial Intelligence (AI), ongoing research continues to discover new forms of intelligence that are efficient and closely mimic an organism’s brain in terms of performance. This article presents a new concept termed wet-neuromorphic computing, in which biological cells or organisms are leveraged to perform computational tasks using their natural molecular functions. We map key neuromorphic properties to natural biological computing observed in bacteria, 3D organoids, and C. elegans. To expand beyond the inspiration of the brain to create conventional neuromorphic computing, the study presents a case study that demonstrates bacterial AI computing using the Gene Regulatory Neural Network (GRNN) derived from E. coli’s gene regulatory network for pattern recognition, validated through wet lab experiments. Lastly, challenges and future directions are discussed.

With the COVID-19 pandemic, the world is confronting various healthcare issues, and healthcare automation is more crucial than ever. The pandemic has revealed the limitations of existing digital healthcare systems to manage public health emergencies. There is no registered population for many healthcare institutions in Sri Lanka, as a result, there is a communication gap. Electronic Health Record systems (EHRs) are becoming popular to share patient details but accessing scattered data across several EHRs while safeguarding patient privacy remains a challenge. Most of these medical records are in printed format and manually entering those into EHR systems is time-consuming and error prone. Not only that pharmaceutical error is a critical healthcare problem, but it is even riskier to visit doctors for pharmaceutical diagnosis during a pandemic. This research introduces a Blockchain-based patient health record system, an Optical Character Recognition (OCR) and Natural Language Processing (NLP) based Medical Document Scanner, a Drug Identifier based on Image Processing and a Medical Chatbot powered by NLP as four novel approaches to address these issues. Altogether with the results, this research aims at introducing a solution for the limitations in healthcare while providing a distributed healthcare framework for the healthcare community worldwide.

Automation is one of the best ways to make easier everyone's day-to-day life. When considering the industry, if someone could be able to automate the mechanism of solving computer software issues, that will be helpful to every company. Because most companies are facing difficulties when it comes to giving IT support. Some companies are suffering from a lack of IT supporters because they have to solve the same problem on different computers at the same time. This most likely reduces the efficiency of the company and affects its performance as well. Without the proper knowledge of technology, employees tend to get a lot of technical issues while working with the technology. Here comes the opportunity to automate the IT-related problem-solving mechanism. That will save a lot of time and increase the company's efficiency as well. While carrying out the background research we thought, by automating the IT-related problem it will reduce the allocation of human power and that will directly affect company efficiency. The research team came up with the idea of creating a software product that can export an executable file by capturing selected tasks of the user's screen and its' relevant processes. Task and process capturing toolkit is a software product called the “ClickMe toolkit”. It can capture relevant processes that are happening in any computer which runs on Windows 10 platform, then make a script called “ClickMe script”, and convert it into an executable file. It can be run on any Windows 10 Platform.

Sri Lanka is an agricultural country since ancient times. Today’s agriculture field is in a dangerous situation because farmers are losing their yield. There are many factors to consider when planting crops like rainfall, temperature, soil conditions, future prices, diseases, etc. We decided to help them through the android application we are making. Here we identified four main problems. First, it was wrong crop cultivation. This is the main reason crops and cultivation are destroyed. To give a solution to that problem, we suggest the five most suitable crops to cultivate according to their location. The second problem is a lack of knowledge about future market prices. As a solution to that problem, we predict prices for each cop for the next 12 months. Another problem is an inability to sell their product at a reasonable price. Here, we directly connect buyers and sellers by removing intermediaries. The last problem is the difficulty to identify diseases affected by crops. Using our mobile app farmers can identify which disease affected their crops by uploading an image to the app. To give solutions to the above-mentioned problems Machine Learning algorithms are used like Random Forest, k-means clustering, and Convolution Neural Network algorithms.