Class X AI Pre-Board Exam 2026
Uploaded by
MradulataClass X AI Pre-Board Exam 2026
Uploaded by
MradulataCommon questions
Powered by AIDevelopers must ensure AI models do not exhibit biases, such as those observed when a hiring tool favors male candidates over females; corrections can involve diversifying training data, involving ethicists in development, and implementing fairness algorithms . Transparency involves clear documentation of decision-making processes to allow for accountability . Such ethical frameworks not only protect user rights but also enhance trust in AI technologies.
Splitting data into train and test sets allows for a preliminary evaluation of a model's performance on unseen data, helping to identify overfitting when a model performs well on training data but poorly on test data . This approach encourages the development of models that generalize well, as the model is optimized to perform not just on the data it was trained on but also on new data.
High precision is critical in applications where minimizing false positives is essential, such as in fraud detection, to avoid unwarranted suspicion or customer inconvenience . Conversely, high recall is crucial in medical diagnosis to ensure all potential cases are captured as a precautionary measure, even if it includes false positives. The balance between precision and recall depends on the application's tolerance for false negatives versus false positives.
Labelled image data is crucial for training supervised computer vision models as it provides the necessary ground truth to learn meaningful patterns and make predictions . The advantage lies in accuracy improvements and the model's ability to understand complex visual inputs. However, it requires extensive manual labor for accurate labeling, and the quality is dependent on the label accuracy, which can be subject to human error or bias.
One major challenge in NLP is the ambiguity in human language, such as when a voice assistant misinterprets a simple command, reflecting the difficulty AI models face in understanding context and nuance . This ambiguity can lead to incorrect predictions or misunderstandings, affecting the overall accuracy and reliability of the system. For instance, misinterpretation of commands can significantly disrupt user interactions and reduce trust in AI systems.
Confusion matrices provide a detailed breakdown of model predictions compared to actual outcomes, offering insights into true positives, false positives, true negatives, and false negatives. Metrics derived from confusion matrices (accuracy, precision, recall, and F1 Score) help assess the model's reliability in various aspects . For example, precision focuses on the correctness of positive predictions, while recall assesses the model's ability to identify all relevant instances, guiding improvements in specific areas of performance.
Dataset imbalance, where one class significantly outnumbers others, leads to models that may predict the majority class well but perform poorly on minority classes. This imbalance skews evaluation metrics like accuracy, making them misleading as they don't reflect true performance across all classes . It poses challenges in achieving equitability in predictions, often necessitating advanced techniques like resampling and alternative metrics like F1 Score for balanced assessments.
NLP is fundamental in enabling computers to understand and respond to human language. Key features for effective chatbot communication include speech recognition for processing spoken language, syntax and semantic analysis for understanding context, and sentiment analysis for gauging user emotions. These elements allow for more intuitive and natural interactions between humans and AI systems .
Testing on unseen data is crucial to gauge a model's ability to generalize beyond the specific training set . Three main reasons include: ensuring robustness and preventing overfitting, evaluating the model's predictive performance in real-world scenarios, and identifying any potential biases that were not apparent during training. This practice helps confirm that the model's learning is not limited to the idiosyncrasies of the training data alone.
The evaluation stage is crucial in assessing an AI model's performance by testing it on a separate dataset and ensuring that it generalizes well to new, unseen data . This step identifies issues such as overfitting or underfitting and allows developers to adjust model parameters and strategies to enhance reliability. Without proper evaluation, a model might perform well on training data but fail in real-world scenarios.