diff --git a/FLOWCHARTS/CH 02/Minimizing loss functions with gradient descent/Update weights and bias.png b/FLOWCHARTS/CH 02/Minimizing loss functions with gradient descent/Update weights and bias.png new file mode 100644 index 000000000..3646f9bce Binary files /dev/null and b/FLOWCHARTS/CH 02/Minimizing loss functions with gradient descent/Update weights and bias.png differ diff --git a/FLOWCHARTS/CH 02/Minimizing loss functions with gradient descent/Update weights and bias.svg b/FLOWCHARTS/CH 02/Minimizing loss functions with gradient descent/Update weights and bias.svg new file mode 100644 index 000000000..c094ad0a0 --- /dev/null +++ b/FLOWCHARTS/CH 02/Minimizing loss functions with gradient descent/Update weights and bias.svg @@ -0,0 +1,75 @@ + + + + + + + Compute the weight's partial derivative + + + $\frac{\partial L}{\partial w_j} = \frac{-2}{n} \sum_i{(y^{(i)} - \sigma(z^{(i)}))}x_j^{(i)}$ + + + Compute the bias's partial derivative + + + $\large \frac{\partial L}{\partial b} = \frac{-2}{n} \sum_i{(y^{(i)} - \sigma(z^{(i)}))}$ + + + $\Delta w_j = \eta \frac{\partial L}{\partial w_j}$ + + + $\Delta b = \eta \frac{\partial L}{\partial b}$ + + + Update the weights + + + $ \textbf{w}:=\textbf{w}+\Delta \textbf{w} $ + + + Update the bias + + + $b:=b+\Delta b$ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + false + + 0.9 + 0.497 + 0.403 + + + diff --git a/FLOWCHARTS/CH 02/Perceptron Learning Rule/m.png b/FLOWCHARTS/CH 02/Perceptron Learning Rule/m.png new file mode 100644 index 000000000..77c052fe7 Binary files /dev/null and b/FLOWCHARTS/CH 02/Perceptron Learning Rule/m.png differ diff --git a/FLOWCHARTS/CH 03/Decision Trees/.$Decision Trees.svg.bkp b/FLOWCHARTS/CH 03/Decision Trees/.$Decision Trees.svg.bkp new file mode 100644 index 000000000..2c2f87427 --- /dev/null +++ b/FLOWCHARTS/CH 03/Decision Trees/.$Decision Trees.svg.bkp @@ -0,0 +1,4 @@ + + + +
Get the features list
Get the features list
For every feature : 
Compute the information gain (IG) 
For every feature :...
Select the feature
with the largest IG
Select the feature...
Set the feature as a parent node
Set the feature as a pare...
Split into left subtree and right subtree at the current root
Split \(D_p\) into left subtree \(D...
Yes
Yes
Stopping criterion met?
Stopping criterion met?
Read the original dataset
Read the original dataset \(D_p...
Output tree model
Output tree model
No
No
Yes
Yes
More subtrees exist?
More subtrees exist?
No
NoCompute IG
Compute impurity on parent node , left child , and right child
Compute impurity on parent node...
Get number of training examples at:

parent node , 
left child node ,
 right child node  
Get number of training...
Compute IG:
Compute IG:...
Select the impurity measure function
Select the impurity mea...Text is not SVG - cannot display \ No newline at end of file diff --git a/FLOWCHARTS/CH 03/Decision Trees/Combining multiple decision trees via random forests.png b/FLOWCHARTS/CH 03/Decision Trees/Combining multiple decision trees via random forests.png new file mode 100644 index 000000000..994e919e4 Binary files /dev/null and b/FLOWCHARTS/CH 03/Decision Trees/Combining multiple decision trees via random forests.png differ diff --git a/FLOWCHARTS/CH 03/Decision Trees/Combining multiple decision trees via random forests.svg b/FLOWCHARTS/CH 03/Decision Trees/Combining multiple decision trees via random forests.svg new file mode 100644 index 000000000..f5a3e0252 --- /dev/null +++ b/FLOWCHARTS/CH 03/Decision Trees/Combining multiple decision trees via random forests.svg @@ -0,0 +1,4 @@ + + + +
Define:
the sample size ,
the number of repetitions ,
and the number of selected features 
Define:...
Read the dataset
Read the dataset
Draw a random bootstrap sample of size
Draw a random bootstrap sample...
Randomly select features without replacement 
Randomly select \(d\) features...
Select the feature with maximum information gain
Select the feature with maximu...
Split the current parent node using the selected feature
Split the current parent node...
Repeat times
Repeat \(k\) times
Aggregate predictions using majority vote
Aggregate predictions using ma...Text is not SVG - cannot display \ No newline at end of file diff --git a/FLOWCHARTS/CH 03/Decision Trees/Decision Trees.png b/FLOWCHARTS/CH 03/Decision Trees/Decision Trees.png new file mode 100644 index 000000000..f581d9104 Binary files /dev/null and b/FLOWCHARTS/CH 03/Decision Trees/Decision Trees.png differ diff --git a/FLOWCHARTS/CH 03/Decision Trees/Decision Trees.svg b/FLOWCHARTS/CH 03/Decision Trees/Decision Trees.svg new file mode 100644 index 000000000..8626fbee7 --- /dev/null +++ b/FLOWCHARTS/CH 03/Decision Trees/Decision Trees.svg @@ -0,0 +1,4 @@ + + + +
Get the features list
Get the features list
For every feature : 
Compute the information gain (IG) 
For every feature :...
Select the feature
with the largest IG
Select the feature...
Set the feature as a parent node
Set the feature as a pare...
Split into left subtree and right subtree at the current root
Split \(D_p\) into left subtree \(D...
Yes
Yes
Stopping criterion met?
Stopping criterion met?
Read the original dataset
Read the original dataset \(D_p...
Output tree model
Output tree model
No
No
Yes
Yes
More subtrees exist?
More subtrees exist?
No
NoCompute IG
Compute impurity on parent node , left child , and right child
Compute impurity on parent node...
Get number of training examples at:

parent node , 
left child node ,
 right child node  
Get number of training...
Compute IG:
Compute IG:...
Select the impurity measure function
Select the impurity mea...Text is not SVG - cannot display \ No newline at end of file diff --git a/FLOWCHARTS/CH 03/KNN/KNN.png b/FLOWCHARTS/CH 03/KNN/KNN.png new file mode 100644 index 000000000..21e564ad2 Binary files /dev/null and b/FLOWCHARTS/CH 03/KNN/KNN.png differ diff --git a/FLOWCHARTS/CH 03/KNN/KNN.svg b/FLOWCHARTS/CH 03/KNN/KNN.svg new file mode 100644 index 000000000..56199cbbe --- /dev/null +++ b/FLOWCHARTS/CH 03/KNN/KNN.svg @@ -0,0 +1,4 @@ + + + +
Choose a distance metric
Choose a distance metric
Define a number
Define a number \(k\)
Compute the distance between the new sample and the samples of the dataset
Compute the distance between the new...
Find the k-nearest neighbors
Find the k-nearest neighbors
Assign the class label by majority vote
Assign the class label by majority v...
Euclidean distance: 
Euclidean distance:...
Manhatten distance: 
Manhatten distance:...
Minkowski distance: 
Minkowski distance:...Distance metricsText is not SVG - cannot display \ No newline at end of file diff --git a/FLOWCHARTS/CH 03/Kernel trick for SVM/SVM.png b/FLOWCHARTS/CH 03/Kernel trick for SVM/SVM.png new file mode 100644 index 000000000..aa970c5c7 Binary files /dev/null and b/FLOWCHARTS/CH 03/Kernel trick for SVM/SVM.png differ diff --git a/FLOWCHARTS/CH 03/Kernel trick for SVM/SVM.svg b/FLOWCHARTS/CH 03/Kernel trick for SVM/SVM.svg new file mode 100644 index 000000000..98b88bb84 --- /dev/null +++ b/FLOWCHARTS/CH 03/Kernel trick for SVM/SVM.svg @@ -0,0 +1,4 @@ + + + +
Choose a kernel function
Choose a kernel function
Apply to the train dataset
Apply \(k\) to the train d...
Tune the hyperparameters
Tune the hyperparameters
Train the SVM model
Train the SVM modelTraining
Test dataset
Test dataset
Apply
Apply \(k\)
Model 
Model 
Weights
Weights
Output
OutputEvaluation Text is not SVG - cannot display \ No newline at end of file diff --git a/FLOWCHARTS/CH 03/Logistic Regression/Logistic Regresion.svg b/FLOWCHARTS/CH 03/Logistic Regression/Logistic Regresion.svg new file mode 100644 index 000000000..2653f3755 --- /dev/null +++ b/FLOWCHARTS/CH 03/Logistic Regression/Logistic Regresion.svg @@ -0,0 +1,4 @@ + + + +

Compute the net input
Compute the net input...

Update the weights and bias
Update the weights and bias...

Compute the loss
Compute the loss...Training

Apply the Sigmoid activation
Apply the Sigmoid activation...

Compute the error
Compute the error...

Compute the net input
Compute the net input...

Apply the Sigmoid activation
Apply the Sigmoid activation...
?
\(\sigma(z) \geq 0.5\) ?
Return 1
Return 1
Yes
Yes
Return 0
Return 0
No
NoEvaluation
Model (weights and bias)
Model (weights and bias)Text is not SVG - cannot display \ No newline at end of file diff --git a/FLOWCHARTS/CH 03/Logistic Regression/Logistic Regression.png b/FLOWCHARTS/CH 03/Logistic Regression/Logistic Regression.png new file mode 100644 index 000000000..227476d1c Binary files /dev/null and b/FLOWCHARTS/CH 03/Logistic Regression/Logistic Regression.png differ diff --git a/FLOWCHARTS/CH 03/Steps of supervised ML algorithm/Supervised ML algorithm.png b/FLOWCHARTS/CH 03/Steps of supervised ML algorithm/Supervised ML algorithm.png new file mode 100644 index 000000000..8adf309e7 Binary files /dev/null and b/FLOWCHARTS/CH 03/Steps of supervised ML algorithm/Supervised ML algorithm.png differ diff --git a/FLOWCHARTS/CH 03/Steps of supervised ML algorithm/Supervised ML algorithm.svg b/FLOWCHARTS/CH 03/Steps of supervised ML algorithm/Supervised ML algorithm.svg new file mode 100644 index 000000000..6607d24d3 --- /dev/null +++ b/FLOWCHARTS/CH 03/Steps of supervised ML algorithm/Supervised ML algorithm.svg @@ -0,0 +1,4 @@ + + + +
Collect labeled data
Collect labeled data
Select features
Select features
Choose a performance metric
Choose a performance...
Choose the model
Choose the model
Evaluate the model
Evaluate the model
Tune the model
Tune the modelText is not SVG - cannot display \ No newline at end of file