Q1) Case Study Question: "Analyze a real-world case study where deep learning was applied for image

classification. In your discussion, cover the following aspects:

1. Problem Statement: What was the specific image classification problem being addressed?

2. Dataset and Preprocessing: What dataset was used, and how was it preprocessed before
feeding into the deep learning model?

3. Model Architecture: What deep learning model architecture (e.g., CNN, VGG, ResNet) was
selected, and why was this architecture suitable for the problem?

4. Training Process: How was the model trained, and what optimization techniques (e.g., data
augmentation, dropout, transfer learning) were employed to enhance performance?

5. Results and Evaluation: What were the key results of the model, and how was its performance
evaluated using metrics like accuracy, precision, recall, or F1 score?

6. Challenges and Improvements: What challenges were encountered during the case study, and
how could the model be improved in future iterations?"

Q2) Case Study Question: "Examine a real-world case study that integrates the use of ImageNet for
object detection and WaveNet for audio generation or analysis. In your response, address the following
components:

1. Problem Definition: Describe the specific problem addressed by integrating ImageNet for
object detection and WaveNet for audio processing (e.g., multimedia analysis, autonomous
systems).

2. Dataset: How was the ImageNet dataset used for object detection, and what type of audio
data was processed using WaveNet? Describe the preprocessing steps for both image and
audio data.

3. Model Architecture & Design: Discuss the architecture used for object detection (e.g., Faster
R-CNN, YOLO) and how WaveNet was implemented for audio analysis or generation. Why were
these models chosen for their respective tasks?

4. Training and Optimization: How were the models trained, and what optimization techniques
were applied (e.g., transfer learning, data augmentation, hyperparameter tuning)?

5. Results and Performance Metrics: Analyze the performance of both object detection and
audio models using appropriate evaluation metrics (e.g., mAP for detection, audio quality
measures).

6. Challenges and Future Improvements: What were the main challenges faced in integrating
ImageNet for object detection with WaveNet for audio processing? Suggest potential
improvements for the models or future applications for such an integration."
Q3) Case Study Question: "Explore a real-world case study where sentiment analysis was applied to
analyze textual data. In your analysis, address the following aspects:

1. Problem Statement: What was the specific sentiment analysis problem being addressed (e.g.,
customer feedback analysis, social media sentiment monitoring)?

2. Dataset: What dataset was used (e.g., reviews, tweets, survey responses), and how was it
prepared and cleaned for sentiment analysis?

3. Model Selection: Which machine learning or deep learning models were used for sentiment
analysis (e.g., Naive Bayes, SVM, LSTM, BERT), and why was the model chosen?

4. Feature Engineering & Preprocessing: How were features extracted from the text (e.g., TF-
IDF, word embeddings, tokenization)? What preprocessing steps (e.g., stop-word removal,
lemmatization) were necessary to prepare the text for modeling?

5. Training and Evaluation: How was the model trained, and what metrics (e.g., accuracy,
precision, recall, F1 score) were used to evaluate its performance?

6. Challenges and Improvements: What challenges were faced during the sentiment analysis
process (e.g., handling sarcasm, class imbalance), and how could the model be improved in
future iterations?"

Q4) Case Study Question: "Analyze a real-world case study where text preprocessing was a critical
component in building a natural language processing (NLP) system. In your analysis, cover the following
aspects:

1. Problem Statement: What was the NLP task that required text preprocessing (e.g., text
classification, sentiment analysis, machine translation)?

2. Dataset: Describe the dataset used (e.g., product reviews, social media posts, news articles).
What were the common challenges encountered in the raw text data (e.g., noise, misspellings,
mixed languages)?

3. Preprocessing Techniques: What specific preprocessing steps were applied (e.g., tokenization,
stop-word removal, stemming/lemmatization, lowercasing)? Why were these steps chosen for
the given task?

4. Feature Representation: How was the processed text represented for the model (e.g., Bag of
Words, TF-IDF, word embeddings)? Discuss why this representation was effective for the case
study.

5. Challenges and Solutions: What challenges were encountered during text preprocessing (e.g.,
handling special characters, dealing with domain-specific language), and how were they
addressed?

6. Impact on Model Performance: How did the preprocessing steps improve the performance of
the NLP model? Use performance metrics (e.g., accuracy, F1 score, precision, recall) to
highlight the improvements."
Q5) Case Study Question: "Discuss a real-world case study where a chatbot was developed
to solve a specific business or customer service problem. In your analysis, address the
following aspects:
1. Problem Definition: What was the specific problem the chatbot aimed to address (e.g.,
customer support, e-commerce, healthcare advice)? Why was a chatbot chosen as the
solution?
2. Dataset: What data was used to train the chatbot (e.g., customer queries, product FAQs,
medical consultations)? Describe the steps taken to clean and preprocess this data.
3. Chatbot Architecture: What type of chatbot was implemented (e.g., rule-based,
retrieval-based, generative)? What models or frameworks were used (e.g., RNNs,
Transformers, GPT, Dialogflow)?
4. Training and Deployment: How was the chatbot trained and optimized for the given
task (e.g., intent recognition, dialogue management, natural language generation)?
Discuss any transfer learning or fine-tuning applied.
5. Evaluation and Metrics: How was the chatbot’s performance evaluated (e.g., user
satisfaction, intent accuracy, response relevance)? What metrics were used to measure
its effectiveness?
6. Challenges and Future Improvements: What challenges were faced during the
development of the chatbot (e.g., understanding complex queries, handling multi-turn
dialogues)? Suggest ways in which the chatbot could be improved in the future."