Tulshihar Bhagawan Patil

@bvuniversity.edu.in

Assistant Professor
Bharati Vidyapeeth (Deemed to be University), College of Engineering, Pune-43

Tulshihar Bhagawan Patil


            Download Compact PDF  
https://researchid.co/tbpatil

RESEARCH, TEACHING, or OTHER INTERESTS

Computer Engineering, Computer Science, Computer Networks and Communications, Computer Science Applications
10

Scopus Publications

Scopus Publications

  • Intelligent Code Optimization: A Knowledge-Based Approach to Error Reduction
    Patil Tulshihar Bhagawan, D. Joshi Shashank, Jitendra Rajpurohit, Milind Gayakwad, Aditi Sharma, Anand Shinde
    Lecture Notes in Networks and Systems, 2026
  • Personalized Learning Systems Using Machine Learning for Adaptive Education
    Monali Gulhane, Utkarsha Aniket Gode, Trupti Chandak, Piyush Pal, Surendra Pal Singh, Tulshihar B. Patil
    2026 International Conference on Emerging Smart Computing and Informatics Esci 2026, 2026
  • Applying the transfer learning models on the dataset on the effect of diseases on Nagvel-betel (Piper betle) leaves
    Milind Gayakwad, Rahul Joshi, Tulshihar Patil, Pratvina Talele, Gurunath S. Waghale, Rajendra Pawar, Nidhi Poonia, Sachin Kadam, Priyanka Paygude
    Data in Brief, 2025
    The dataset of Betel leaves includes 4156 leaves affected by various diseases. These diseases include Leaf Spot, Powdery Mildew, Anthracnose, Bacterial Blight, Cercospora Leaf Spot, Sooty Mold, Downy Mildew, Wilt Disease, Rust Disease, Mosaic Virus, Black Rot, Root Rot, Stem Canker, Leaf Curl Disease, and Fusarium Wilt. The camera is used to collect high-resolution images to ensure the exact detection of the images to detect diseases. The resolution of the photos was 3000 × 4000, consuming approximately 3 mb. The data set covers a wide range of diseases, and many samples were collected under each category. The dataset is saved using a hierarchical data structure, as the name of the folder indicates the label or category of the image. The reuse and recreation of this type of dataset are ensured by mapping the name of the disease with the apparent characters of the disease on the leaves. The experiment was performed using Vision Transformer Models to check the robustness of the dataset. The result of the classification report states that the range of accuracy varies from 0.7 to 0.9.
  • Learning style prediction of e-learner using hybrid optimizer-based neural network
    Snehal Rathi, Priyanka Paygude, Nazim Shaikh, Supriya Sawant-Patil, Tulshihar Patil, et al.
    Journal of Integrated Science and Technology, 2025
    The Learning Style prediction model in e-learning systems has gained immense attention in the area of education. In the current scenario, the major demand for online platforms is to provide a substantiated interface that acclimatizes the learning styles of the learners. People learn in different ways, and their preferences can change over time. The accurate prediction of learning style can raise the learners' learning gain. This Research proposed a technique to predict the learning styles, by capturing the interaction behavior of the learner. The learning styles are predicted and grounded on the uprooted features using a Neural Network. It is trained and classified using a hybrid optimizer which is a fusion of Squirrel Search (SS) and Rider Optimization Algorithm (ROA). Felder-Silverman Learning Style Model is used to map the learner's learning styles. Eventually, the pupil and course ID, learning style, course completion status, and test score data are recorded to find the correlation. The proposed hybrid optimizer-based model provides superior performance compared to techniques with an accuracy of 0.95 and a maximal correlation of 0.406.
  • Statistical Signal Processing for Radar Systems
    Rajesh Kedarnath Navandar
    Panamerican Mathematical Journal, 2025
    Through measurable models, "Factual Flag Handling for Radar Frameworks" goes into detail around the progressed procedures and strategies utilized to see at and progress radar information. This field employments thoughts from both measurable hypothesis and flag preparing to unravel difficult issues in radar framework execution, like finding targets, taking after them, and speculating what they are. This article talks approximately distinctive factual models and strategies that can be utilized with radar. These incorporate Bayesian gauges, versatile sifting, and theory testing. This paper talks around how factual apparatuses can be utilized to create radar frameworks more precise, dependable, and clear by mimicking and decreasing the impacts of commotion, impedances, and outside components. Creating factual flag preparing methods for lessening disarray, distinguishing targets, and estimating parameters are a few of the foremost critical subjects. Other imperative points incorporate ways to analyze and get it radar information in genuine time. It gives a careful see at both the hypothetical bases and real-world applications, appearing how measurable strategies can be utilized to make strides radar framework capabilities. There are case thinks about and real-life examples that appear how these strategies can be utilized totally different radar circumstances, such as military, flying machine, car, and climate radars. This book gives engineers and understudies the instruments they got to move forward radar framework plan and execution in settings that are getting more complicated and changing rapidly by combining factual flag preparing with radar innovation.
  • Leveraging real-time data: A location-based ambulance booking and tracking system with geofencing
    Prashant Chavan, Priyanka Paygude, Snehal Rathi, Mahesh M. Patil, Tulshihar Patil, et al.
    Journal of Integrated Science and Technology, 2025
    Delayed ambulance arrival is a critical issue in emergency medical care, impacting patient outcomes. This research addresses the challenge of slow response times in India's ambulance services. Our proposed system aims to significantly improve response times and patient transportation efficiency. To provide rapid medical assistance, the technology allows dynamic ambulance reservations depending on the severity of the accident, deploys competent drivers, and transmits real-time traffic data. The paper details the system's architecture and approach, including the use of Firebase Authentication and AWS Amplify for user management and backend development. Integration of various APIs (Google Maps API, Google Places Auto Complete, and Distance Matrix API) enhances user experience and functionality. This research presents a real-time ambulance booking and tracking system with the potential to revolutionize emergency medical response in India.
  • Real-Time Clickstream Data Processing and Visualization Using Apache Tools
    T. B. Patil, Keshav Anand, Aditya Bhateja, Kashif Jamal, S. T. Sawant-Patil, Priyanka Paygude
    2023 7th International Conference on Computing Communication Control and Automation Iccubea 2023, 2023
    In the modern digital generation, clickstream data analytics is a crucial component of online selling and purchasing platforms. It offers useful information for enhancing user experience, conversion rates, website design, and keeping track of the success of digital marketing efforts. A data warehouse is required to store and analyze clickstream data, which requires a lengthy and complex ETL process. Two of the leading large data processing tools and distributed streaming systems are Spark and Kafka. This paper presents a comprehensive examination of recent research pertaining to clickstream data analytics, real-time data visualization, data lifecycle management utilizing Apache technologies. Based on the review, the paper proposes an architecture that can aid organizations and local merchants in effectively visualizing the performance of their products. This paper provides a valuable resource for researchers and practitioners seeking to explore the latest developments in their future work.
  • Enhancing image security by employing blowfish algorithm further embedding text and stitching the RGB components of a host image
    Deepanshu Agarwal, Pravesh Panwar, Purva Vyas, Tejas Patil, S. D. Joshi, et al.
    International Journal of Recent Technology and Engineering, 2019
    ‘Privacy, privacy everywhere but not a safety method to implement it’: a harsh reality of today’s world. With the precipitation of more data (2 x 1019 bits of data is created in every 86400 seconds) in computer networks, involvement of meta-data in the form of images is essential. To keep data safe and secure in order to inculcate privacy, to eradicate any kind of eavesdropping, and to maintain confidentiality, integrity and availability of it, certain security measures are needed to make in account for. So in order to make it available, we required a technique through which we can securely transfer any kind of data over a network. In practise the information security can be achieved either by using Cryptography or Steganography. The process described in this paper is not a mediocre it is more scrupulous towards the security because it involves image encryption, steganography and image stitching. Initially we are encrypting an image using Blowfish algorithm then we are embedding the secret text into this encrypted image by modifying the least significant bit (LSB) of the image by our data. Moreover, to enhance the privacy and security we are stitching the above resultant image with the red, green and blue (RGB) components of a host image and thereby producing an image more secure than the one which the existing systems can form for data transmission..
  • Monitoring and training stock prediction system for historical & live dataset using Lstm & Cnn
    Omveer Singh Deora, , Pawan Jha, Prof. S.T. Sawant Patil, Prof. T.B. Patil, Dr. S. D. Joshi, , , , and
    International Journal of Innovative Technology and Exploring Engineering, 2019
    A country development and stability are directly associated with its economy and today’s economy is profoundly dependent on the Stock market. Stock market indexes are subject to continuous change with respect to time, a hype or fall in the stock market has a crucial role in deciding the investor’s profit. Due to the economical ups & down and rapid growth in profit from the stock market, there required a need of developing a software application which continuously monitor the stock index’s and a prediction algorithm which can predict the possible change in stock index as for where it can go in future. Prediction of stock market does not follow any rules or predefined guidelines, hence prediction of stock market is difficult to achieve and the data-set for stock market prediction is also non-linear in nature which requires an efficient approach to resolve the time-series dependency of non-linear data. In our proposed system we are using the LSTM (long short-term memory) for efficiently predicting the stock index on historical data and the sudden change in stock market due to number of un-controllable factors is analysed by CNN model. As per the noise in the data-set we are employing wavelet denoising technique. If any changes in stock index with more than 10% of its initial value is analysed by monitoring module, then the system will notify the user with the change and also aggregating the result of predicting algorithm on that specific stock. Using our model Moneypred the accuracy in stock prediction is more than 70%.
  • Big Data Privacy Using Fully Homomorphic Non-Deterministic Encryption
    Tejashree B. Patil, Girish Kumar Patnaik, Ashish T. Bhole
    Proceedings 7th IEEE International Advanced Computing Conference Iacc 2017, 2017
    Big data is a large amount of digital information. Now days, data security is a challenging issue that touches several areas along with computers and communication. The security of data which stored online has become a main concern. Several attackers play with confidentiality of the user. Cryptography is a approach that provide data security to the user. Despite of huge efforts to protect sensitive data, hackers typically manage to steal it. Computing with encrypted data is strategies for safeguarding confidential data. The partial homomorphic encryption is specialized for only one operation on the encrypted data. For example the Pailliers encryption scheme performs only one mathematical operation on encrypted numerical data and is successful to compute the sum of encrypted values. The Pailliers encryption scheme is unable to do multiple mathematical operations on encrypted numerical data. The proposed encryption algorithm computes more than one mathematical operation on encrypted numerical data thereby further protecting the encrypted sensitive information.