The aging population is confronted with an increasingly pressing challenge concerning their mental health and well-being. As individuals advance in age, they become more susceptible to various mental disorders, which, if left undiagnosed and untreated, can significantly impact their quality of life and strain healthcare systems.
Schizophrenia is not just a diagnosis; it's a journey that demands understanding and compassion. One can see various symptoms like : hallucinations, delusions, impaired attention, change in behaviour and more , along with diagnosis using psychiatric evaluation and clinical interviews. Here we strive to arm individuals with knowledge about schizophrenia. By doing so, we aspire to minimize the social stigma surrounding mental health, encouraging open dialogue and nurturing acceptance.
Our mission is clear: to create awareness, foster empathy, and extend a helping hand to those impacted by this complex mental health issue.
Our goal in this endeavor is to enlighten minds and spark conversations. Just as a single spark can light up a dark room, a single piece of knowledge can dispel ignorance. Through this project, we aim to raise public consciousness about the realities of schizophrenia, its risk factors, and its symptoms. Despite its prevalence, many are unaware of the signs that can signal the onset of this condition. By sharing insights and information, we hope to empower individuals to recognize the signs, seek help, and foster supportive environments.
Schizophrenia prediction happens by applying two models we have developed by training two datasets using the Scikit-learn library patched with intel oneDAL extension in Python. This is a code implementation that aims to develop a predictive model for predicting if a person is prone to schizophrenia or not. The code uses the Logistic regression classification technique which was found to be most accurate after comparing 7 different machine learning(classification) algorithms like decision tree, k-nearest neighbor, Gaussian naive Bayes, multinomial naive Bayes, support vector classifier and random forest to predict the likelihood of being prone to schizophrenia based on a range of variables.
The first dataset used in the code includes various columns such as name, gender, age, marital status, fatigue, restlessness, pain, changes in hygiene, and changes in movement scores. By analyzing these(symptoms) variables and using machine learning algorithms to identify patterns and correlations, the predictive models can provide accurate assessments of a patient's risk of being prone to schizophrenia.
The other dataset used in the code includes text and labels, where the text has been preprocessed using NLP algorithms then we trained a model to predict if the text spoken by the person shows fine or poor mental health condition.
The pickle files generated by training the models on Intel Developer Cloud servers and Intel oneAPI kernel for jupyter notebook have been integrated on the deployment server with UI. The prediction result from the models is sent as a prompt to chat-GPT API which produces a sophisticated report. The report has been saved on InterPlanetary File System where it is safeguarded using Blockchain technologies.
- Artificial Intelligence(Natural Language Processing)
- Machine Learning
- Blockchain
- Web Technologies
- Predictions for Schizophrenia
- Generating a summarized report
- Privacy and Security
- Works with multiple languages
- User-friendly interface
🌟 Employ Intel oneDAL for Enhanced Outcomes and Swifter Calculations (Intel oneAPI Data Analytics Library) on Intel Developer Cloud
Since, we used IPFS in local machine we cloudn’t able to deploy the same, but still we had hosted the entire UI and features part!!
✅Leveraging Intel oneDAL for Application Development: Harnessing the capabilities of the Intel oneAPI Data Analytics Library (oneDAL), our project is dedicated to accelerating comprehensive data analytics across the spectrum – from data preprocessing and transformation to analysis, modeling, validation, and decision-making. By optimizing algorithms for various processing modes, oneDAL enhances data ingestion, computational efficiency, and scalability.
✅Crafting a Schizophrenia Susceptibility Prediction Application involves a meticulous journey of research and development. Along this path, I have cultivated a rich tapestry of insights and skills.
✅Mastery of Machine Learning Techniques: A deep comprehension of diverse machine learning algorithms was achieved, with a specific focus on their relevance in predicting an individual's vulnerability to schizophrenia.
✅Data Analysis Proficiency: Hands-on experience was gained in aggregating and analyzing expansive datasets to train machine learning models customized for predicting susceptibility to schizophrenia.
✅Synergistic Collaborations: The realization of this project necessitated seamless teamwork across a multidisciplinary spectrum, encompassing machine learning, data analysis, and beyond. This collaborative endeavor underscored the pivotal role of collective effort in realizing shared aspirations.
✅Integration of Natural Language Processing (NLP): Our project’s innovation extends to the incorporation of NLP techniques, further enriching the predictive model's capabilities.
These instances collectively signify the proficiencies and aptitudes that were nurtured during the development of this predictive model. In essence, the creation of a Schizophrenia Susceptibility Prediction Application is a rewarding challenge that necessitates a fusion of technical prowess and comprehensive knowledge.
58764a2cd1f82ee13494efc64de9a42effbc9c95