From 39951d053090d933a5f6520710993fa9361b9b73 Mon Sep 17 00:00:00 2001 From: Christian Kolset Date: Fri, 5 Dec 2025 17:19:09 -0700 Subject: Added module 5 lectures in markdown --- README.md | 13 +++++++++++-- .../convolutional_neural_networks_for_engineering_images.md | 10 ++++++++++ tutorials/module_5/data_for_engineering.md | 11 +++++++++++ tutorials/module_5/failure_and_fatigue_prediction.md | 10 ++++++++++ tutorials/module_5/how_neural_networks_learn.md | 7 +++++++ tutorials/module_5/introduction_to_ai.md | 13 +++++++++++++ tutorials/module_5/neural_networks_core_concepts.md | 8 ++++++++ tutorials/module_5/supervised_learning_techniques.md | 13 +++++++++++++ ...ised_learning_clustering_and_dimensionality_reduction.md | 10 ++++++++++ 9 files changed, 93 insertions(+), 2 deletions(-) create mode 100644 tutorials/module_5/convolutional_neural_networks_for_engineering_images.md create mode 100644 tutorials/module_5/data_for_engineering.md create mode 100644 tutorials/module_5/failure_and_fatigue_prediction.md create mode 100644 tutorials/module_5/how_neural_networks_learn.md create mode 100644 tutorials/module_5/introduction_to_ai.md create mode 100644 tutorials/module_5/neural_networks_core_concepts.md create mode 100644 tutorials/module_5/supervised_learning_techniques.md create mode 100644 tutorials/module_5/unsupervised_learning_clustering_and_dimensionality_reduction.md diff --git a/README.md b/README.md index fcba973..23f85f7 100644 --- a/README.md +++ b/README.md @@ -84,5 +84,14 @@ See the tutorials under the /tutorials/ directory or use the links below. - [Data Visualization and Presentation](tutorials/module_4/4.7_Data_Visualization_and_Presentation.md) #### Module 5: Introduction to AI for Engineering Problems -- [[supervised_vs_unsupervised_learning]] -- [[neural_networks]] \ No newline at end of file +- [Introduction to AI](tutorials/module_4/introduction_to_ai.md) +- [Neural Network Core Concepts](tutorials/module_4/neural_networks_core_concepts.md) +- [How Neural Networks Learn](tutorials/module_4/how_neural_networks_learn.md) +- [Supervised learning Techniques](tutorials/module_4/supervised_learning_techniques.md) +- [Data for Engineering](tutorials/module_4/data_for_engineering.md) +- [Unsupervised learning: Clustering & Dimensionality Reduction](tutorials/module_4/unsupervised_learning_clustering_and_dimensionality_reduction.md) +- [Convolutional Neural Network for Engineering Images](tutorials/module_4/convolutional_neural_networks_for_engineering_images.md) +- [Failure and Fatigue Prediction](tutorials/module_4/failure_and_fatigue_prediction.md) + +## Resources +- [Projects/Assignments/Problems](practical_problems.md) \ No newline at end of file diff --git a/tutorials/module_5/convolutional_neural_networks_for_engineering_images.md b/tutorials/module_5/convolutional_neural_networks_for_engineering_images.md new file mode 100644 index 0000000..81df8b5 --- /dev/null +++ b/tutorials/module_5/convolutional_neural_networks_for_engineering_images.md @@ -0,0 +1,10 @@ +Lecture 7: Convolutional Neural Networks for Engineering Images + +- Why fully connected NNs fail on images +- Convolution: filters, feature maps +- Pooling & downsampling +- CNN pipeline overview +- Applications: + - Crack detection in metals/concrete + - manufacturing defect identification + - thermal-image pattern recognition \ No newline at end of file diff --git a/tutorials/module_5/data_for_engineering.md b/tutorials/module_5/data_for_engineering.md new file mode 100644 index 0000000..0b8ef6f --- /dev/null +++ b/tutorials/module_5/data_for_engineering.md @@ -0,0 +1,11 @@ +Lecture 5: Data for Engineering: Features, Labels, Normalization, Metrics + +- Train/validaiton/test splits +- Normalization & scaling +- Features engineering basics +- Metrics: accuracy, MSE, ROC, F1 +- Engineering-specific constraints: + - sensor noise + - unit consistency + - small datasets + - measurement bias \ No newline at end of file diff --git a/tutorials/module_5/failure_and_fatigue_prediction.md b/tutorials/module_5/failure_and_fatigue_prediction.md new file mode 100644 index 0000000..c4ac70b --- /dev/null +++ b/tutorials/module_5/failure_and_fatigue_prediction.md @@ -0,0 +1,10 @@ +Lecture 8: AI for Failure & Fatigue Prediction + +- Using time-series sensor data +- Features: peaks, cycles, RMS, stress/strain sequences +- Classical models vs ANN-based predictors +- Introduction to RUL (Remaining Useful Life) concepts +- Real-world challenges: + - limited failure examples + - imbalanced data + - saftey requirements and false positives \ No newline at end of file diff --git a/tutorials/module_5/how_neural_networks_learn.md b/tutorials/module_5/how_neural_networks_learn.md new file mode 100644 index 0000000..f81ec4d --- /dev/null +++ b/tutorials/module_5/how_neural_networks_learn.md @@ -0,0 +1,7 @@ +Lecture 3: How Neural Networks Learn + +- Loss functions +- Gradient descent & optimization +- Backpropagation concept +- Underfitting, overfitting, regularization +- Hyperparametersupervised Learning: Regression for engineering systems \ No newline at end of file diff --git a/tutorials/module_5/introduction_to_ai.md b/tutorials/module_5/introduction_to_ai.md new file mode 100644 index 0000000..17477c5 --- /dev/null +++ b/tutorials/module_5/introduction_to_ai.md @@ -0,0 +1,13 @@ +Lecture 1: Introduction to AI + +- Historical context: Expert systems -> machine learning -> deep learning +- What is AI? Deep learning? Neural Networks? +- What makes AI useful in engineering: automation, prediction, anomaly detection +- Types of engineering problems where ai is superior. +- Application: + - Predictive maintenance + - Quality control + - CFD/FEA surrogate models +- Limitations, misconceptions and ethical considerations +- Short demo: Show a simple trained model making prediction (linear regression) + diff --git a/tutorials/module_5/neural_networks_core_concepts.md b/tutorials/module_5/neural_networks_core_concepts.md new file mode 100644 index 0000000..4c81e91 --- /dev/null +++ b/tutorials/module_5/neural_networks_core_concepts.md @@ -0,0 +1,8 @@ +Lecture 2: Neural Networks: Core Concepts + +- Perceptron, weights, bias +- Multi-layer perceptron (MLP) +- Activation functions +- Forward pass (conceptual math only) +- Why neural networks can learn nonlinear engineering behavior +- Small example: predicting a nonlinear spring force curve \ No newline at end of file diff --git a/tutorials/module_5/supervised_learning_techniques.md b/tutorials/module_5/supervised_learning_techniques.md new file mode 100644 index 0000000..c84155a --- /dev/null +++ b/tutorials/module_5/supervised_learning_techniques.md @@ -0,0 +1,13 @@ +Lecture 4: Supervised Learning Techniques + +- Definition: supervised learning = labeled data +- Regression tasks +- Classification tasks +- Classical ML models + - linear & logistic regression + - decision trees + - k-NN +- Engineering examples + - Predicting tool wear + - classifying vibration patterns + diff --git a/tutorials/module_5/unsupervised_learning_clustering_and_dimensionality_reduction.md b/tutorials/module_5/unsupervised_learning_clustering_and_dimensionality_reduction.md new file mode 100644 index 0000000..3f704b5 --- /dev/null +++ b/tutorials/module_5/unsupervised_learning_clustering_and_dimensionality_reduction.md @@ -0,0 +1,10 @@ +Lecture 6: Unsupervised learning: Clustering & Dimensionality Reduciton + +- Definition: +- k-means and hierachical clsutering +- PCA for dimensionality reduction +- Applications: + - Grouping failure modes + - anomaly detection in vibration data + - clustering material micro structure patterns + -- cgit v1.2.3