Algorithm 3 Training algorithm

1.  Construct set from sensor data: Given training set $S$, test set $T$, feature dimension $F$.   Determine the parameters: The number of decision trees $t$, the depth of each tree $d$, and the number of features $f$ used by each node.   Termination conditions: The minimum number of samples on the node $S$, the minimum information gain on the node $m$. 2. From $S$, there is a training set $S\left(i\right)$ of the same size as the extracted size $S$, as a sample of the root node, and training is started from the root node. 3. If: the termination condition is reached on the current node,    Set the current node as a leaf node.   If: the current node does not reach the termination condition,    The $f$ dimensional features are randomly selected from the $F$ dimensional features without replacement. Using this $f$ dimensional feature, find the best one-dimensional feature $k$ and its threshold $th$.    The sample whose $k$ dimension feature is less than $th$ at the current node is divided into left nodes, and the rest is divided into right nodes.    Continue to train other nodes. 4. Repeat 2, 3 until all nodes have been trained or marked as leaf nodes. 5. Repeat 2, 3, 4 until all decision trees have been trained.