iPhone Premium Analysis: Hardware vs. Price
Regression models trained on 1,359 smartphones show that Apple devices are priced above their hardware-justified value.
A regression study of whether Apple devices are priced above what their hardware alone would justify. I trained models on 1,359 smartphones to predict price from hardware specifications, then compared the predictions against actual iPhone prices.
Approach
I tested linear regression, decision trees, random forests, and a multilayer perceptron in Python (pandas, scikit-learn). The MLP generalised best, reaching R² = 0.94 on unseen devices, which made it the most trustworthy yardstick for “what should this phone cost given its hardware?”
Finding
Flagship iPhones are priced significantly above their hardware-predicted value, while older models trend back toward parity. The premium is concentrated in the newest devices.