-
logo
Why ASDQuickScreen Should Win?
ASDQuickScreen is a pioneering project that excels in several categories, making it a strong contender for the following awards:
Best Design Project: ASDQuickScreen showcases exceptional design innovation by creating a user-friendly and accessible tool for Autism Spectrum Disorder (ASD) assessment. The project's thoughtful design ensures that users, including healthcare professionals and individuals, can easily navigate the screening process from the comfort of their own homes. This design elegance streamlines the initial assessment of ASD, making it both efficient and effective.
Best Social Impact: ASDQuickScreen has an undeniable social impact by addressing a critical need in the field of healthcare. Early detection and intervention for ASD are key factors in improving the lives of individuals on the autism spectrum. By providing an accessible, cost-effective, and accurate screening tool, ASDQuickScreen empowers parents, caregivers, and professionals to take proactive steps toward diagnosis and intervention. The project's potential to positively impact lives and communities is substantial.
Best Use of IBM zSystems: ASDQuickScreen utilizes IBM zSystems to enhance the project's scalability, security, and reliability. By harnessing the power of IBM zSystems, we ensure that user data is secure and that the screening process remains robust and efficient. The utilization of IBM zSystems aligns perfectly with the project's mission to provide a dependable and accessible tool for ASD assessment.
Best Project: ASDQuickScreen stands out as a testament to innovation, social impact, and effective use of technology. The project has leveraged machine learning and IBM zSystems to create a game-changing tool in the field of healthcare. By enhancing accessibility, affordability, and accuracy in ASD screening, ASDQuickScreen embodies the spirit of the "Best Project" category, showcasing excellence in both execution and impact.
In summary, ASDQuickScreen's innovative design, significant social impact, effective use of IBM zSystems, and overall project excellence position it as a strong candidate for recognition in the specified categories. It embodies the spirit of innovation, technology, and positive societal change.
Inspiration
The inspiration behind this project lies in addressing the limitations of existing Autism Spectrum Disorder (ASD) screening tools. These tools are often costly, time-consuming, and less accessible to individuals who may benefit from early ASD assessment. We aimed to create an innovative, cost-effective, and user-friendly solution that empowers both healthcare professionals and individuals to conduct an initial ASD screening quickly and conveniently.
What it does
ASDQuickScreen serves as a preliminary screening tool that can be used from the comfort of one's home, either on a computer or a mobile device. It offers a swift and efficient assessment of ASD, helping users determine if further formal diagnostic procedures are necessary. By employing data preprocessing techniques and robust evaluation methods, this model provides accurate predictions and empowers users to make informed decisions.
How we built it
The project began with data preprocessing, which involved cleaning and eliminating irrelevant fields. To enhance the data's suitability for modeling, we used label encoding to transform non-numerical labels into numerical ones, and we normalized the data. The models were evaluated using a 10-fold cross-validation technique, ensuring robust and consistent performance. This approach generated accuracy scores and confusion matrices to assess the model's predictive capabilities.
Challenges we ran into
During the project, we encountered challenges in data preprocessing, specifically in identifying which attributes were truly relevant for prediction. This required a thoughtful and thorough analysis of the data to refine our model effectively.
Accomplishments that we're proud of
One of the project's notable achievements is the high accuracy achieved by our models, with averages ranging from 90.7% to 97.4%. These results are complemented by low standard deviations, underscoring the reliability and consistency of our model's performance.
What we learned
Through this project, we learned that certain attributes, such as gender and age, have less influence on predicting ASD. Conversely, factors like jaundice at birth, family history, ethnicity, and behavioral traits are vital contributors to accurate ASD screening. Additionally, we gained insights into the importance of optimizing feature selection for enhanced model performance.
What's next for ASDQuickScreen
Looking ahead, the project's future entails expanding the dataset by collecting more information. This will enable us to further refine and improve the accuracy of the model. The ongoing goal is to provide an even more accessible and reliable tool for ASD screening, benefiting individuals and healthcare professionals alike in their quest for early ASD detection and support.
Log in or sign up for Devpost to join the conversation.