SDSujata Deshmukh

Sujata Deshmukh

*This is an algorithmically generated author record based on publicly available information.

Sujata Deshmukh has a total of 70 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Contributors

    by Yogesh Ahirwar, Neha Sharad Bhagwat, Pankaj Chandre, Sujata Deshmukh, Pankaj A. Dhakate, Nitin Dhawas, Vaishali Nitin Dhawas, Bharti Dhote, S. Ganesh, Mayuri Gawade, et al.

  • Future trends in AI-enhanced digital twins: From autonomous systems to quantum integration

    by Vaishali Sarbhukan, Anandkumar Pardeshi, Sujata Deshmukh

  • Interpretable AI for Skin Cancer Prediction using Clinical Indicators for Indian Patients

    by Tushar Jagannath Surwadkar, Sujata Deshmukh

    Skin cancer is a serious public health challenge, and early diagnosis is crucial for effective treatment. We present a machine-learning pipeline to classify patients as “Cancer” vs “Non-Cancer” using clinical and demographic features. The study uses a real-world dataset ($\mathbf{n}=\mathbf{7 0 4}$) with 20 predictors (age, gender, lesion site, lifestyle factors, etc.) and an extremely imbalanced target (696 non-cancer vs 8 cancer cases), fortunately, due to less cancer patients observed in India. We perform thorough preprocessing (including handling missing values and categorical encoding), exploratory data analysis, and compare five classifiers: Logistic Regression, Random Forest, Gradient Boosting, Support Vector Machine (RBF kernel), and K-Nearest Neighbors. A stratified 5-fold cross-validation ensures reliable model comparison and overfitting control. The best model (Logistic Regression) achieved near-perfect cross-validated performance (ROC-AUC $\boldsymbol{\approx} \mathbf{1. 0}$) and very high test accuracy $(99.3 \%)$, with precision 66.7% and recall 100% on the rare cancer class. Performance metrics (accuracy, precision, recall, $\text{F1}$, ROC-AUC), confusion matrix, ROC curve, and feature importance (from the tree-based models) are reported. We include visualizations from EDA (class balance, age distribution, lesion site distribution, correlation heatmap) and evaluation (cross-validation bar chart, confusion matrix, ROC curve, and feature importances). The study highlights the predictive power of routine clinical features and provides insights for interpretable skin cancer risk assessment using SHAP and LIME.

  • Open Access

    DishaDoot: Career navigation platform

    by Vijay Maruti Shelake, Royce Robert Dmello, Jason James Dsouza, Jatin Jaywant Kadu, Sujata Deshmukh

    DishaDoot: Career navigation platform is an innovative software solution designed to guide students from varied social and economic backgrounds in making informed career decisions, moving beyond the traditional and often restrictive approach of following popular career trends or succumbing to societal pressures. The platform aims to address the high unemployment rate in India by encouraging diversification of talent and knowledge, ultimately helping students identify career paths that align with their individual strengths, academic achievements, and personal interests. DishaDoot integrates a multi-faceted recommendation system that considers a variety of factors to provide customized career guidance. This includes user-profile test generation to assess personality traits and skills, career recommendations based on academic performance (such as 10th and 12th grade percentages), aptitude test scores, hobbies, and other personal interests. Additionally, DishaDoot introduces a scholarship recommendation module that suggests financial aid options based on the selected course, aiding students in navigating their educational pursuits more affordably. The career recommendation system’s training dataset was curated by scraping data from various sources, focusing specifically on undergraduate courses offered by the University of Mumbai to streamline the scope of the project. By providing accurate and diverse course recommendations. DishaDoot is more than just a career recommendation system; it’s a mission-driven platform aligned with the United Nations sustainable development goal 4, which advocates for inclusive, equitable, and quality education for all. Rooted in this global vision, DishaDoot aims to bridge the gap between students’ potential and the opportunities available to them. In today’s competitive world, students often face confusion, lack of guidance, and limited access to resources when it comes to planning their careers. DishaDoot steps in as a reliable companion on their journey, offering more than just course suggestions. One of the platform’s standout features is its newly integrated scholarship recommendation engine, which helps students discover financial aid opportunities tailored to their academic background and interests, something that can make a real difference for those from underprivileged backgrounds. To further support students, DishaDoot also includes a smart support bot that provides real-time assistance, answers common queries, and ensures a smoother, more personalized user experience. These thoughtful additions address the practical challenges students face, from exploring career paths to securing resources for their education. What sets DishaDoot apart is its holistic approach. It not only suggests career paths based on the students’ interests, academic performance, and aptitude, but also guides them with tools and insights that make the process more informed and empowering. It’s designed with the Indian education ecosystem in mind, helping students especially those from rural or less-connected regions make confident, well-rounded career choices. DishaDoot isn’t just a platform; it’s a step towards transforming how young minds in India navigate their futures. By combining technology with empathy and purpose, DishaDoot aspires to create a meaningful impact on the lives of students, opening doors to opportunities and helping them turn aspirations into achievements.


  • Optimization to release net zero emission in built environment from newly developed natural fiber–hybrid composite

    by Tamilselvan Amuthan, Nagaraj Nagaprasad, Vijayan Somurajan, Nakul Gupta, Ramakrishnan Kulasekaran Sathish Kumar, Ramaswamy Krishnaraj, Vijilius Helena Raj, R. Srinivas, Kuldeep K. Saxena, Sujata Deshmukh

    This article focuses on investigating the dynamic behavior of hybrid composites involving natural fibers. The objective of this study is to examine and evaluate the effectiveness of recently used hybrid composite materials. The primary materials utilized are glass fiber-reinforced plastic. Plastic is applied to the surface of fiber-reinforced epoxy composites. Natural fibers are also used in other materials, polyester-based hybrid composites. The breakage pressure of the double shear test in the first composition has been improved from 1.10 to 1.38[ ]kN. The hardness rating was enhanced from 76.15 to 87.99. The interlamination experimental findings improved from 7.96 to 8.52[ ]kN. The results showed that a hybrid with a 50:50 comparable strand ratio of glass and sisal outscored a hybrid with a small fraction of sisal fiber. The second composite describes the effects of natural filler loading at 5%, 10%, and 15% on the mechanical behavior of a polyester-based hybrid. Materials created in a reputable environment for automotive and marine applications. Both hybrid nanocomposite preparation methods are tested, and the findings are validated. The electrical and mechanical characteristics of the materials are also validated and demonstrated to be superior to other current composites. This study highlights composite materials and technological advances and discusses natural fibers’ rich history and promise in many applications.

  • Modeling and simulation through multi-criteria approach for customized conveyor-type industrial oven handling wide variety of metal structures

    by Dipali Bhole, Sujata P. Deshmukh, Sujata Kolhe

    Modern ovens are utilized from large span of time for completing different heat treatment cycles like baking, curing, annealing, aging, drying, and so on. There are numerous sorts of industrial ovens namely batch, relieving, drying, baking, reflow, clean room, conveyor types, and so on. Conveyor oven helps to perform the heat treatment cycle in an uninterrupted flow, including the heating of the objects followed by the cooling down of the same object. A multipurpose conveyor-type industrial oven is designed using analytical approach for heat treatment process on various objects. This article presents the modeling and CFD analysis of conveyor-type continuous ovens. Simulation of the entire model of conveyor-type industrial oven is carried out in Solidworks Flow package. Since the boundary conditions of an actual industrial oven is open at both entry and exit points, same condition is maintained in the CFD analysis. With the help of the simulation model, researchers can study and observe the temperature distribution throughout the oven.