Impara Visualizing Predictive Results | Predictive People Analytics

Scorri per mostrare il menu

Visualizing the results of predictive models is essential in people analytics because it bridges the gap between complex data science and actionable HR decisions. Stakeholders such as HR managers, executives, and team leads may not be familiar with technical model outputs or statistical jargon. By presenting predictions visually, you make it easier for these decision-makers to grasp key insights, understand risks, and take informed actions. Visualization transforms numbers into stories that resonate and drive engagement.


              1234567891011121314151617181920
            
import matplotlib.pyplot as plt
import numpy as np

# Example employee IDs and their predicted attrition probabilities
employee_ids = ['E101', 'E102', 'E103', 'E104', 'E105']
predicted_probs = [0.15, 0.62, 0.45, 0.80, 0.30]

plt.figure(figsize=(8, 4))
bars = plt.bar(employee_ids, predicted_probs, color='skyblue')
plt.ylim(0, 1)
plt.xlabel('Employee ID')
plt.ylabel('Predicted Attrition Probability')
plt.title('Predicted Attrition Probabilities for Employees')

# Highlight employees with high attrition risk (probability > 0.5)
for bar, prob in zip(bars, predicted_probs):
    if prob > 0.5:
        bar.set_color('salmon')

plt.show()

Probability plots like the one above help you quickly identify employees at higher risk of leaving. The y-axis shows the likelihood of attrition, with higher bars indicating greater risk. When communicating these results, emphasize that the model predicts probability, not certainty. Use clear visuals and color coding to draw attention to high-risk individuals, and be ready to explain what factors might contribute to their risk. This approach enables HR stakeholders to prioritize interventions and allocate resources effectively.


              1234567891011121314151617181920212223
            
import matplotlib.pyplot as plt
import numpy as np

# Example actual and predicted outcomes (1 = left, 0 = stayed)
employee_ids = ['E101', 'E102', 'E103', 'E104', 'E105']
actual = [0, 1, 0, 1, 0]
predicted = [0, 1, 1, 1, 0]

x = np.arange(len(employee_ids))
width = 0.35

fig, ax = plt.subplots(figsize=(8, 4))
rects1 = ax.bar(x - width/2, actual, width, label='Actual', color='lightgreen')
rects2 = ax.bar(x + width/2, predicted, width, label='Predicted', color='cornflowerblue')

ax.set_xlabel('Employee ID')
ax.set_ylabel('Attrition (1 = Left, 0 = Stayed)')
ax.set_title('Predicted vs. Actual Attrition Outcomes')
ax.set_xticks(x)
ax.set_xticklabels(employee_ids)
ax.legend()

plt.show()

1. Why visualize predictive model results in HR?

2. Fill in the blank: Probability plots help communicate ____ to stakeholders.

3. What can a bar chart of predicted vs. actual outcomes reveal?

Tutto è chiaro?

Grazie per i tuoi commenti!

Sezione 3. Capitolo 4

Chieda ad AI

Chieda pure quello che desidera o provi una delle domande suggerite per iniziare la nostra conversazione

Sezione 3. Capitolo 4