Questo sito utilizza cookie per raccogliere dati statistici.
Privacy Policy
# Overfitting vs Underfitting
![][image12]
L'**underfitting** e l'**overfitting** sono due concetti fondamentali nell'ambito del **Machine Learning** e si riferiscono al comportamento del modello durante il processo di apprendimento. Entrambi riguardano la capacità del modello di generalizzare correttamente i dati, ossia di fare previsioni accurate su nuovi dati non visti.
---
## Underfitting
L'**underfitting** si verifica quando un modello è **troppo semplice** per rappresentare i dati correttamente. In altre parole, non ha abbastanza capacità per catturare le relazioni significative tra le variabili nei dati di addestramento. Questo porta il modello a commettere errori sia nei dati di addestramento che nei dati di test, e generalmente ha una bassa accuratezza.
### Cause dell'underfitting:
- Il modello è troppo semplice (ad esempio, un modello lineare applicato a dati non lineari).
- Non ci sono abbastanza caratteristiche (feature) rilevanti.
- Il numero di parametri del modello è troppo piccolo.
- Il tempo di addestramento è troppo breve.
### Esempio visivo:
Immagina di voler approssimare i dati distribuiti in una curva complessa (non lineare) con una semplice retta. Il modello non riuscirà a catturare le variazioni dei dati, e la retta sarà una rappresentazione inadeguata.
### Conseguenze:
- Il modello ha **un'alta varianza** sugli errori, ossia commette molti errori perché non riesce a modellare correttamente i dati.
- Scarso rendimento sia sui dati di addestramento che sui dati di test.
### Soluzione:
- Utilizzare un modello più complesso.
- Aggiungere più caratteristiche rilevanti (feature engineering).
- Aumentare il tempo di addestramento o migliorare l'algoritmo di apprendimento.
---
## Overfitting
L'**overfitting** si verifica quando un modello è **troppo complesso** e si adatta troppo bene ai dati di addestramento, catturando anche il rumore e le caratteristiche specifiche dei dati che non sono generalizzabili a nuovi dati. Questo porta il modello a performare bene sui dati di addestramento, ma male sui dati di test, poiché ha "imparato" dettagli non rilevanti che non si ripresenteranno in nuovi dati.
### Cause dell'overfitting:
- Il modello ha troppi parametri rispetto ai dati (ad esempio, un modello polinomiale di grado molto alto).
- Il modello è stato addestrato troppo a lungo sui dati di addestramento.
- I dati di addestramento contengono troppo rumore o outlier, e il modello cerca di adattarsi a essi.
- Troppi dettagli sono stati inclusi nelle caratteristiche del modello.
### Esempio visivo:
Immagina di voler approssimare i dati distribuiti in una curva semplice (ad esempio, una linea o una curva regolare), ma utilizzi un modello che si adatta esattamente a tutti i punti (inclusi gli outlier). Otterrai una funzione molto complessa e tortuosa, che si adatta perfettamente ai dati di addestramento ma fallisce miseramente quando incontra nuovi dati.
### Conseguenze:
- Il modello ha una **bassa generalizzazione** e performa male su nuovi dati, pur avendo un'ottima accuratezza sui dati di addestramento.
- Performante sui dati di addestramento, ma scadente sui dati di test.
### Soluzione:
- **Ridurre la complessità del modello** (ad esempio, utilizzando meno parametri o un grado polinomiale più basso).
- **Utilizzare tecniche di regolarizzazione** (come L1 o L2) per penalizzare la complessità del modello.
- Aumentare la dimensione del set di dati di addestramento.
- **Early stopping**: interrompere l'addestramento prima che il modello si adatti troppo ai dati di addestramento.
- Applicare tecniche di cross-validation per valutare meglio le prestazioni del modello su set di dati diversi.
---
## Relazione tra Underfitting, Overfitting e Complessità del Modello
L'underfitting e l'overfitting possono essere visualizzati come due estremi opposti lungo un continuum di complessità del modello:
- **Modelli semplici** (come i modelli lineari per dati complessi) rischiano di underfittare perché non riescono a catturare le caratteristiche complesse dei dati.
- **Modelli complessi** (come le reti neurali molto profonde o i modelli polinomiali di alto grado) rischiano di overfittare, perché si adattano troppo bene ai dettagli specifici dei dati di addestramento, incluso il rumore.
Il grafico seguente potrebbe illustrare la relazione:
```
|--------------|--------------|--------------|
Underfitting Modello Ideale Overfitting
```
Il **modello ideale** si trova nel mezzo, dove il modello è abbastanza complesso da catturare le relazioni nei dati, ma non troppo complesso da adattarsi al rumore.
---
## Come individuare underfitting e overfitting
Per determinare se un modello soffre di underfitting o overfitting, si può confrontare l'accuratezza o l'errore sui dati di addestramento e sui dati di test:
- **Underfitting**: scarsi risultati sia sui dati di addestramento che sui dati di test.
- **Overfitting**: buoni risultati sui dati di addestramento, ma scarsi sui dati di test.
### Esempio pratico:
Immaginiamo di costruire un modello di classificazione su un dataset.
| Modello | Accuratezza Addestramento | Accuratezza Test |
|----------------|---------------------------|------------------|
| Modello A | 50% | 48% | --> Underfitting
| Modello B | 95% | 85% | --> Buon Modello
| Modello C | 99% | 60% | --> Overfitting
---
## Conclusione
- **Underfitting**: il modello non ha appreso a sufficienza dai dati e ha prestazioni scarse sia sui dati di addestramento che sui dati di test.
- **Overfitting**: il modello ha imparato troppo dai dati di addestramento, adattandosi anche al rumore, e performa bene sui dati di addestramento ma male su nuovi dati.
- La chiave per un buon modello di Machine Learning è trovare il giusto equilibrio tra complessità e capacità di generalizzazione, evitando sia l'underfitting che l'overfitting.
[image12]: <data:image/png;base64,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>