ACCT 331: WEEK 9

These flashcards cover key concepts from the ACCT 331 lecture on applied AI, focusing on significant partnerships, metrics for classification models, the importance of model interpretability, and strategies for effective AI implementation.

What partnership did Walmart establish to enhance their shopping experience?

Walmart partnered with OpenAI to create AI-first shopping experiences.

How will Walmart's AI-first shopping experience benefit customers?

It allows customers to complete purchases directly within ChatGPT, making the retail experience more proactive.

What does the AI Research Associate Program at JPMorgan Chase aim to explore?

The program aims to advance research in AI, including machine learning and cryptography, for the benefit of JPMorgan's clients.

What is predicted about AI companies according to Byron Deeter?

The next major wave of successful companies will emerge in consumer-facing AI applications.

What is the primary focus for companies looking to use AI in regard to data?

Companies want personnel who can explain data implications and guide business decisions.

What does the term 'confusion matrix' refer to in model evaluation?

A confusion matrix is a table that visualizes the performance of a classification model by displaying the correct and incorrect predictions.

What is the precision metric in classification models?

Precision measures the proportion of correct positive predictions to the total predicted positives.

What is recall in the context of classification metrics?

Recall measures the proportion of actual positives that were correctly identified.

What is the F1-Score?

The F1-Score is the harmonic mean of precision and recall, used to balance the trade-offs between these two metrics.

Why can accuracy be misleading in classification tasks?

Accuracy can be misleading, especially in imbalanced datasets where the model may achieve high accuracy by only predicting the majority class.

What does the ROC curve illustrate?

The ROC curve plots the true positive rate against the false positive rate at various threshold settings for a classifier.

What is model interpretability?

Model interpretability refers to how easily humans can understand the reasoning behind the predictions made by a model.

What is a major challenge with complex models like deep learning?

They are often considered 'black boxes' and can be difficult to interpret or explain.

What is model bias, and how can it occur?

Model bias arises when models are trained on historical data that reflect societal biases, leading to discriminatory outcomes.

What are some factors that contribute to model fairness?

Fair models must avoid discrimination based on sensitive characteristics, such as race or gender.

What does OOS stand for regarding inventory and sales predictions?

OOS stands for 'Out Of Stock,' referring to products that are not available for sale.

How can K-means clustering be used in inventory management?

K-means clustering groups similar transactions to analyze typical basket compositions and customer purchasing patterns.

What is the purpose of feature engineering in predictive modeling?

Feature engineering involves creating new features from raw data to improve model performance.

Why is a comprehensive exploratory data analysis important before modeling?

EDA helps identify critical data patterns and informs feature engineering decisions to prevent costly errors in modeling.

What is the significance of 'average recency' in predicting customer behavior?

Average recency captures customer purchase frequency and loyalty patterns, making it a strong predictor of sales.