Sagar Suraj Lachure

RESEARCH, TEACHING, or OTHER INTERESTS

Computer Science, Artificial Intelligence, Computer Science Applications, Environmental Engineering

Scopus Publications

Advanced deep learning approaches for superior temporal analysis and forecasting of water level discharge for the Bamni River
S. S. Lachure and Ashish Tiwari
Informa UK Limited



Statistical Analysis of Flood-Drought Trend in Central India and the West-Coast
S. S. Lachure, J. S. Lachure, A. D. Sawarkar, K. R. Singh, S. Sahu, A. Lohidasan, and N. D. Dhamele
Springer Nature Singapore



A Lightweight Hybrid Quantum Convolution Neural Network for Temperature Forecasting
Sagar Lachure, Lalit Damahe, Jaykumar Lachure, Ankush Sawarkar, Swaraj Singh Bhati, Rishi Chhabra, and Nikita Dhamele
AIP Publishing



Enhancing Environmental Resilience: Precision in Air Quality Monitoring through AI- Driven Real- Time Systems
Ankit Mahule, Kaushik Roy, Ankush D. Sawarkar, and Sagar Lachure
CRC Press



Quantum machine learning applications to address climate change: A short review
Sagar Suraj Lachure, Ashwin Lohidasan, Ashish Tiwari, Meera Dhabu, and Neeraj Dhanraj Bokde
IGI Global
In the previous three to four decades, numerical weather and climate modelling have advanced significantly, yet many challenges still exist. Appropriate adaptation techniques to reduce loss of life and property require geographically and temporally targeted predictions. Artificial Intelligence and machine learning (AI and ML) based technologies are improving predictions. However, they are bound by the absence of a hardware's or a software's—or both—capable of handling the enormous data volumes created on a global basis. The burgeoning paradigm of quantum computing (QC) has potential applications across many industries. This review shows that the current progress in quantum ML for quantum computers may lead to technological advancements in climate change research. The subsequent climate forecasting improvements are expected to have several socioeconomic benefits. The authors have also provided three or four examples showing how quantum technology might be used with ML systems to study climate change.


Commercial Indian Bamboo Species Classification on matK DNA Barcode Sequences using Machine Learning Techniques with K-mer
Ankush D. Sawarkar, Deepti D. Shrimankar, Lal Singh, Anurag Agrahari, Sagar Lachure, and Neeraj Dhanraj Bokde
IEEE
Bamboo, a grass, belongs to the Poaceae family, with 1642 species from 116 genera worldwide. It has exceptional physical, chemical, and mechanical qualities, which allow it to be employed in over a thousand different ways and contribute to a trade value of USD 2.76 billion. Bamboo is grown using rhizomes, tissue culture, or short branch cuttings without any other checks resulting in incorrect species identification and categorisation. Therefore, the classification or identification of these bamboo use its DNA barcode sequences with a K-mer based method, and machine learning (ML) is the most excellent strategy for resolving issues with the conventional or traditional categorisation of the species. A DNA barcode is a brief genetic signature that helps identify the species to which an organism belongs. It is possible to extract a useful feature from genome sequences using K-mer based approaches, which may then be used to increase comparison accuracy. In this research, we evaluate the classification performance of four supervised ML models on the DNA-barcode sequence of six Indian commercial bamboo species with a different K-mer combination. For this classification, we choose matK barcode region and supervised ML models such as Logistic Regression (LR), Random Forest (RF), Support Vector Machine (SVM) and Gradient Boosting Machine (GBM). The results analysis of these models on the matK DNA sequence with different K-mers demonstrates that the classification capabilities of the GBM approaches are quite promising, and it has an accuracy of 95.3% on average.


Cloud Engineering-based on Machine Learning Model for SQL Injection Attack
Kavita Singh, Sakshi Kokardekar, Gunjan Khonde, Prajakta Dekate, Nishita Badkas, and Sagar Lachure
IEEE
The entire IT sector has been altered by Cloud Engineering, which is also continually coming up with new, creative solutions to everyday issues. A systematic method for the commoditization, standardization, and governance of cloud computing applications is provided by cloud engineering, which is the application of the engineering discipline to cloud computing. These days, a lot of businesses and individual users keep enormous amounts of data on the cloud. The information in the data could be sensitive or could be pertinent to the client or the user. Through a variety of harmful methods, the attackers attempt to obtain private data. One such attack is a SQL Injection Attack, in which a hostile user attempts to get unauthorized access to sensitive cloud data by injecting malicious SQL queries. In this paper, a machine learning model is proposed to identify this kind of attack. The Count-Vectorizer approach and the TF-IDF Vectorizer approach are used to test 3 different machine learning models on two different datasets, one with 4200 data items and the other with 30100 data items, both of which contain SQL queries. The models were built using the Random Forest, XgBoost, and Extra Trees Classifier algorithms, respectively. After deploying the models to the test, it was found that with the test size 0.2, Extra trees classifier shows the highest accuracy on dataset of 4200 entries on both the approaches TF-IDF Vectorizer (TV) and Count Vectorizer (CV) and for the dataset of 30100 entries the XgBoost Classifier gives high accuracy with Count Vectorizer approach while XgBoost gives highest accuracy with TF-IDF Vectorizer approach.


Performance of 125 watt PV module using MATLAB-simulink
Umesh P. Pagrut, A. S. Sindekar, Sagar S. Lachure, and Jaykumar S. Lachure
IEEE
The paper predicts the behavior of PV (Photovoltaic) module by simulation in MATLAB. The study is done on the module of 125 Wp/12Volt/SN80. In the course of study, selected module is observed for its behavioral characteristics at different temperature levels and incident sun radiations at different levels. The model is developed for the different environmental factors on the photovoltaic power generation. Simulink technique is used to get PV module characteristics. The non-linear characteristics curves obtained from it are used to compute parametric values which are compared to the datasheet of manufacturer. The effect of parasitic resistance is observed. The impact of shunt resistance (Rsh) and series resistance (Rs) are studied on the performance of PV module. The observations of studied module may prove to be the guidelines for large module system i.e. arrays and for large scale solar photovoltaic energy conversion system.


Diabetic Retinopathy using morphological operations and machine learning
Jaykumar Lachure, A.V. Deorankar, Sagar Lachure, Swati Gupta, and Romit Jadhav
IEEE
Diabetic Retinopathy that is DR which is a eye disease that affect retina and further later at severe stage it lead to vision loss. Early detection of DR is helpful to improve the screening of patient to prevent further damage. Retinal micro-aneurysms, haemorrhages, exudates and cotton wool spots are kind of major abnormality to find the Non- Proliferative Diabetic Retinopathy (NPDR) and Proliferative Diabetic Retinopathy (PDR). The main objective of our proposed work is to detect retinal micro-aneurysms and exudates for automatic screening of DR using Support Vector Machine (SVM) and KNN classifier. To develop this proposed system, a detection of red and bright lesions in digital fundus photographs is needed. Micro-aneurysms are the first clinical sign of DR and it appear small red dots on retinal fundus images. To detect retinal micro-aneurysms, retinal fundus images are taken from Messidor, DB-rect dataset. After pre-processing, morphological operations are performed to find micro-aneurysms and then features are get extracted such as GLCM and Structural features for classification. In order to classify the normal and DR images, different classes must be represented using relevant and significant features. SVM gives better performance over KNN classifier.


Review on precision agriculture using wireless sensor network