Exp 8 — LSTM weather prediction

3) Dataset

• Source: Use a public weather dataset (e.g., NOAA / Open-Meteo exports / Kaggle) and cite it.
• Features/Labels: [Describe X and y; mention classes if classification]
• Split: [Train/Validation/Test or K-fold]
• Preprocessing: [Scaling/Normalization, resizing, tokenization, etc.]

4) Procedure / Steps

1. Load dataset and perform preprocessing.
2. Define model architecture and justify key choices.
3. Compile model (loss + optimizer + metrics).
4. Train with validation and log curves.
5. Evaluate on test set and compute required metrics.
6. Summarize observations and limitations.

5) Code (Skeleton)

Paste your complete runnable code below (or attach notebook link in the final submission).

import numpy as np
import pandas as pd
import tensorflow as tf
from tensorflow import keras

# TODO: load weather series/features
# TODO: create sequences

model = keras.Sequential([
  keras.layers.Input(shape=(LOOKBACK, FEATURES)),
  keras.layers.LSTM(64),
  keras.layers.Dense(1)
])
model.compile(optimizer='adam', loss='mse', metrics=['mae'])

6) Results / Output

• Metrics: [Write your final values: accuracy/F1 or MAE/MSE]
• Plots: [Attach loss/metric curves; prediction vs actual plots if forecasting]
• Screenshots: [Paste screenshots of outputs, confusion matrix, sample predictions]

7) Observations / Discussion

• [Observation 1: what changed when you tuned epochs/batch size?]
• [Observation 2: evidence of overfitting/underfitting?]
• [Observation 3: what improved performance (augmentation, regularization, fine-tuning)?]

8) Conclusion

Write 3–6 lines summarizing what you implemented, key result, and what you learned.

9) Viva Questions

1. How is LSTM different from a simple RNN?
2. What is the vanishing gradient problem in RNNs?
3. What is the difference between training and validation data?
4. Explain overfitting and two ways to reduce it.
5. Why do we normalize/scale inputs?
6. What does batch size and epoch mean?
7. How do you choose a loss function for a task?

10) Post-lab Assignment

• Try multivariate input (temp, humidity, pressure).
• Tune lookback length and compare.