Reading Guide: Artificial Intelligence and Sustainable Computing (ICSISCET 2023)

Artificial Intelligence and Sustainable Computing: Proceedings of ICSISCET 2023 brings together 50 papers from the 5th International Conference on Sustainable and Innovative Solutions for Current Challenges in Engineering and Technology, held at Madhav Institute of Technology & Science (MITS) in Gwalior, India, on October 21–22, 2023. It is edited by Manjaree Pandit, M. K. Gaur and Sandeep Kumar and published by Springer in the Algorithms for Intelligent Systems series. If you have ever opened a proceedings volume and wondered where to begin, this guide gives you a plan.

Get the book: View Artificial Intelligence and Sustainable Computing (ICSISCET 2023) on Amazon

What this book is (and who it's for)

Like other conference proceedings, this book is a collection of independent research papers rather than a single argument. Each paper stands on its own: a problem, a proposed approach, experiments and conclusions. According to the publisher, the research centres on AI methods that blend nature-inspired algorithms, evolutionary computing, fuzzy systems, computational intelligence, machine learning and deep learning. The editors pitch these methods as useful for real problems because they learn from data, adapt easily and do not need a precise model of the system.

The publisher presents the book as a reference for both beginners and experienced researchers. In practice, it works best for:

• Students choosing a thesis or capstone topic who want to see what problems are being worked on right now.
• Researchers writing a literature review on applied soft computing or AI for engineering.
• Engineers and technical managers scanning for methods they could pilot in their own systems.

A working knowledge of basic statistics and machine learning concepts will help. You do not need to be an expert in every technique; you will pick up a lot by comparing how different teams apply them.

Before you start

• Decide what you want out of it. A literature review, a project idea and a practical method to try at work each call for a different reading strategy (see below).
• Set up a comparison template. For each paper, note the problem, method, data source, evaluation metric, baseline and headline result.
• Prepare a sandbox. If you plan to reproduce anything, a Python environment with scikit-learn and a deep learning framework will cover most applied machine learning work. Optimization libraries for genetic algorithms or particle swarm methods are useful too.
• Refresh the core terms. Evolutionary algorithms, fuzzy inference systems, convolutional and recurrent neural networks, and common metrics such as accuracy, F1 score and RMSE come up often in this kind of research.

How to read it

The volume is presented as a single list of papers, so the best approach is to build your own path. Start by reading every title in the table of contents and tagging each paper with one of the themes below.

Theme A: Optimization inspired by nature

Swarm, evolutionary and other population-based methods applied to design, scheduling or resource allocation. Engineers in energy, power and manufacturing should begin here.

Theme B: Reasoning under uncertainty

Fuzzy logic and related computational intelligence methods. These are valuable when inputs are noisy or human judgment needs to be encoded. Control and decision-support readers will find this group most relevant.

Theme C: Learning from data

Machine learning and deep learning applied to prediction, classification and detection. This is the best starting point for students who want a hands-on project.

Theme D: Sustainability in practice

Papers that tie AI to efficiency, resource savings or environmental goals. Read these if your interest is in the "sustainable computing" half of the title.

Three reading plans

• The literature reviewer: read the abstract, method and conclusion of every paper in one theme. Build a table and look for gaps no paper addresses.
• The project seeker: pick two or three Theme C papers with public datasets. Read them fully, then try to reproduce one result before extending it.
• The practitioner: scan Themes A and D for problems that resemble yours. Read the experimental setup closely to judge whether the conditions match your real-world constraints.

On a first pass, skim long hyperparameter tables and detailed derivations. Come back to them only for the papers you decide to build on.

Key ideas to take away

• Flexible methods for messy problems. The book's central case is that adaptive, learning-based techniques suit engineering problems that resist exact modelling.
• Combining methods often pays off. Applied AI research frequently pairs techniques, such as using optimization to tune a learning model.
• Read results critically. Check the baseline, the dataset size and whether results were tested on unseen data before trusting a headline number.
• Research moves fast. Comparing this volume with the previous year's proceedings shows how quickly applied AI priorities shift.

Practice and discussion questions

1. Pick one paper and explain its problem and method in three sentences to a non-specialist. What did you have to leave out?
2. For an optimization paper, what does the method improve on, and by how much? Is the difference practically meaningful?
3. Which papers use a public dataset? How would that affect your confidence in their results?
4. Compare one fuzzy-logic paper with one deep learning paper. When would you choose each approach?
5. Where does a paper claim a sustainability benefit? What evidence would convince you it holds in real deployment?
6. Identify one open question the authors raise in their conclusions. How might you design a follow-up study?
7. If you also read the ICSISCET 2022 volume, which themes grew or faded between the two years?

Pair it with

• The previous edition. Artificial Intelligence and Sustainable Computing: Proceedings of ICSISCET 2022, by the same editors and also in our store, gives you a second snapshot for comparison.
• Metaheuristics and optimization basics. Understanding why swarm and evolutionary methods work makes it much easier to judge the papers that use them.
• Machine learning evaluation. Cross-validation, baselines and statistical significance are the tools you need to read applied results with confidence.

Start with one question

The best way into a proceedings volume is a clear question and a small stack of well-chosen papers. Use the themes above, take good notes, and you will come away with methods and ideas you can apply. Get Artificial Intelligence and Sustainable Computing (ICSISCET 2023) on Amazon, or explore more titles on the BooksPlusAi shelf.

As an Amazon Associate, BooksPlusAi may earn from qualifying purchases.

Sources

• Springer: Artificial Intelligence and Sustainable Computing, Proceedings of ICSISCET 2023
• e-bookshelf listing for the ICSISCET 2023 proceedings
• Amazon product page