Random Forests

# Random Forests
### Dr. D’Agostino McGowan

---

<div class="my-footer">
<span>
Dr. Lucy D'Agostino McGowan <i>adapted from slides by Hastie & Tibshirani</i>
</span>
</div>

---

## The heart disease example

_Recall that we are predicting whether a patient has heart disease from 13 predictors_
---

## 1. Randomly divide the data in half, 149 training observations, 148 testing

```r
set.seed(77)
heart_split <- initial_split(heart, prop = 0.5)
heart_train <- training(heart_split)
```

---

## 2. Create model specification

```r
model_spec <- rand_forest(
  mode = "classification",
  mtry = ---
) %>% 
  set_engine("ranger")
```

---

## 2. Create bagging specification

```r
bagging_spec <- rand_forest(
  mode = "classification",
* mtry = 13
) %>% 
  set_engine("ranger")
```

---

## 2. Create Random Forest specification

```r
rf_spec <- rand_forest(
  mode = "classification",
* mtry = 3
) %>% 
  set_engine("ranger")
```

* The default for `rand_forest` is `floor(sqrt(# predictors))` (so 3 in this case)

---

## 3. Fit the model

```r
model <- fit(rf_spec,
             HD ~ Age + Sex + ChestPain + RestBP + Chol + Fbs + 
               RestECG + MaxHR + ExAng + Oldpeak + Slope + Ca + Thal,               
             data = heart_train)
```

---

## 4. Examine how it looks in the test data

```r
heart_test <- testing(heart_split)
model %>%
  predict(new_data = heart_test) %>%
  bind_cols(heart_test) %>%
  conf_mat(truth = HD, estimate = .pred_class) %>%
  autoplot(type = "heatmap")
```

![](18-random-forest-in-r_files/figure-html/unnamed-chunk-9-1.png)

---

## <i class="fas fa-laptop"></i> `AE 05 - Random Forests`

1. Open your application exercise (05) from last class
2. Refit your model as a _random forest_
3. Knit, Commit, Push. Be sure to have the final results pushed to GitHub by **April 10 at noon**