The Quantum Pharaoh: Rewriting the Future of the Past

Imagine this...You’re standing in the heart of the Egyptian desert.Beneath your feet lies a lost civilization — untouched for thousands of years.Above your head? The cloud of data, algorithms, and… quantum circuits.What if quantum computing could help us decide what to dig up next?What if the mysteries of the past… needed the power of the future to be revealed?

This project is not just about artifacts, it’s about prioritizing cultural survival using Quantum Machine Learning.

Welcome to:“The Quantum Pharaoh” — Where Ancient Egypt meets Qiskit.

Our Team(8) — Small, Passionate & Powerfully Curious

We’re a small team with a big dream: To explore how quantum technology can help us protect Egypt’s ancient heritage.

Each one of us brings something unique — from coding skills to creative thinking — but what unites us is our shared love for both history and innovation.

We laugh, we learn, we debug together — and somehow, between late-night calls and broken code, we built something that just might make a difference.

Because we believe the past deserves a future — powered by AI.

Why This Project?

❝The past is disappearing… faster than we can study it. What if AI — even better, Quantum AI — could help us prioritize what to save?❞

1

Egypt's heritage is at risk: looting, erosion, modern expansion

2

We need intelligent prediction of which sites matter most

3

Let’s try something bold: merging quantum computing with archaeological science

What Makes This Project Unique?

First time applying Quantum Neural Networks to Egyptian archaeological data

Combines classical models & Qiskit QNN to compare power

Built on real archaeological indicators: looting risk, sonar detection, climate impact

Visualized insights + feature rankings + real evaluation metrics

Simulated on CPU — but built for the future of real quantum hardware

How I Did It — Step-by-Step

🛠️ My Hybrid ML Pipeline (Classical + Quantum)

# Load & clean data

df = pd.read_csv("Dataset.csv")df.dropna(inplace=True)

# Encode categorical columns

for col in ['Time Period', 'Material Composition', 'Script Detected']:df[col + '_encoded'] = LabelEncoder().fit_transform(df[col].astype(str))

Then:

🔬 Split: 80% training / 20% testing

🔄 Normalize using StandardScaler

🤖 Classical Models: RandomForest, XGBoost, TabNet, HuberNN

⚛️ Quantum QNN: Built using ZZFeatureMap + RealAmplitudes + EstimatorQNN

The Quantum Circuit we Built

feature_map = ZZFeatureMap(feature_dimension=6)ansatz = RealAmplitudes(num_qubits=6)qc = feature_map.compose(ansatz)qnn = EstimatorQNN(circuit=qc,input_params=feature_map.parameters,weight_params=ansatz.parameters,estimator=Estimator())

Trained using SPSA optimizer on 100 samples

Quantum model simulated entirely on CPU — yet performed strongly

Model Performance Summary:

Random Forest

RMSE: 13.5127

MAE: 11.4291

R2: -0.0575

XGBoost

RMSE: 14.6071 MAE: 12.0879 R2: -0.2357

TabNet

RMSE: 252.9361

MAE: 197.1785

R2: -369.5099

HuberLoss NN

RMSE: 14.6440

MAE: 12.3333

R2: -0.2419

Quantum NN

RMSE: 74.3992

MAE: 73.2282

R2: -31.0564

Results — How Did Each Model Perform?

0.3

Quantum Model R²

Outperformed expectations

0.7

Classical Models R²

Showed best performance

Quantum model outperformed expectations — R² > 0.3

Random Forest

XGBoost

TabNet

HuberLoss NN

Quantum NN

Classical models were best, but Quantum showed real potential!

Visualization: The Eyes of the Model

Don’t just trust the metrics — see the behavior.

Actual vs Predicted (Classical & Quantum)

Feature Importance — What really matters?

(Materials? Time Period? Looting?)

These plots told me what the data hide — and what the model saw.

Why This Matters

1

This isn't a Kaggle project. This is cultural survival.

2

Help archaeologists focus on the most at-risk sites

3

Build national-scale ML tools to manage excavation budgets

4

Digitally preserve heritage before it’s physically lost

5

Use Quantum AI in humanities — not just physics or finance

My Secret Sauce — How I Got Here

Best Practices I Followed

1

Compared 5 different models

2

Used LabelEncoder for categorical encoding

3

Normalized data before quantum training

4

Reduced training sample for QNN (100 samples for 6 qubits max)

5

Evaluated using RMSE, MAE, R²

6

Added visualization + interpretability layer

And I told a story — not just a presentation.

Real-World Value

“This isn’t just a model — it’s a blueprint for preserving Egypt’s history.”

1

Guide archaeologists to high-value sites

2

Identify vulnerable locations for protection

3

Help governments allocate limited resources

4

Show how Quantum AI can support humanities, not just science

What’s Next?>

 I see Quantum AI becoming part of future archaeology

Run QNN on real IBM Q hardware

Use larger data & hybrid circuits

Collaborate with archaeologists for field validation

Publish results — bridge science and cultural studies

 Thank You!Thank you to our professors and mentors — for your time, your support, and for pushing us to grow in such a short time.You didn’t just teach us quantum...You taught us how to collaborate, how to think deeper, and how to become a real team..And thank you to everyone for listening —We hope you enjoyed our project as much as we enjoyed building it together.