Skip to content

Data Science & Analytics

I clean and explore your data with Python, Pandas and NumPy, then turn it into results that make decisions easier, through statistical models and clear visualizations.

In a customer churn analysis on telecom data I compared several methods, from logistic regression to XGBoost; a Random Forest tuned with hyperparameter optimization gave the best result. In the dam water level study I compared ARIMA, SARIMA, ARIMAX and SARIMAX models and showed the effect of seasonality and external variables.

What is included

  • Data cleaning: missing values, outliers, encoding categorical variables
  • Exploratory data analysis and a look at the relationships in it
  • Statistical modeling and time series forecasting
  • Clear charts with Matplotlib and Seaborn
  • A written report of the findings and recommendations

Process

  1. We sharpen the question We agree on which decision the data should help you make.
  2. I prepare the data I sort out missing, wrong and outlying values and get the data ready for analysis.
  3. I analyze and model I study the relationships and trends and, where needed, build a statistical or time series model.
  4. I present the findings I share the results with charts and in plain words, showing what matters and why.

Technologies

Python, Pandas, NumPy, Statsmodels, scikit-learn, XGBoost, Matplotlib, Seaborn

Example projects

Frequently asked questions

My data is in Excel files. Is that enough?

It is enough to start. Excel, CSV or a database table all work; first we look at the structure and quality of the data together.

How are the timeline and the price decided?

By the scope of the work. A short note on what you need is enough to start; once I have looked at how your data is structured, I send you a quote with the timeline and the price.