I will do time series forecasting with arima, sarima, bayesian and ml in r and python
Statistician and Data Scientist, R Studio, Python and Machine Learning
About this Gig
Need time series analysis or forecasting for sales, demand, finance, energy, agriculture or research? I am Analyst Wadud, a statistician and PhD researcher building reproducible R/Python models for business and research.
My services include:
- EDA: trends, STL decomposition, ACF/PACF, ADF/KPSS, seasonality.
- Statistical forecasts: ARIMA, SARIMA, SARIMAX, ETS, Prophet, Holt-Winters.
- Econometrics: VAR, VECM, cointegration, Granger causality, ARCH/GARCH.
- Bayesian methods: BSTS, BVAR, Bayesian ARIMA, state-space, posterior intervals.
- ML forecasting: SVR, Random Forest, XGBoost, LightGBM, CatBoost.
- Deep learning: LSTM, GRU, N-BEATS, N-HiTS, TFT and hybrids.
- Validation: rolling backtests, residuals, MAE, RMSE, MASE.
Outputs: forecasts with intervals, diagnostics, model comparisons, documented code, Jupyter/Quarto reports and optional dashboards.
Tools: R, Python, RStudio, statsmodels, fable, PyMC, Stan, scikit-learn, PyTorch.
Message me your data, frequency, horizon and objective. Packages cover defined tasks; Bayesian, deep learning, multivariate or deployment projects may require a custom offer.
Other Data Science & ML Services I Offer
FAQ
1. What time series exploration, seasonality and stationarity tests can you do?
I analyze trends, cycles, seasonality, change points and outliers using STL, ACF/PACF, differencing, ADF, KPSS and residual diagnostics. I select tests and transformations according to the data and modeling assumptions.
2. Do you provide frequentist ARIMA, SARIMA and classical forecasting?
Yes. I work with AR, MA, ARIMA, SARIMA, ARIMAX/SARIMAX, ETS, Holt-Winters, state-space and seasonal baseline models. I assess suitability through diagnostics, time-ordered evaluation and forecast horizons.
3. Can you perform econometric time series and volatility analysis?
Yes, as applicable: VAR, VECM, ARDL/ECM, cointegration, Granger causality, impulse responses, ARCH/GARCH, structural breaks and volatility forecasting. Complex multivariate econometric studies need a custom quote.
4. Do you provide Bayesian time series models and credible intervals?
Yes, for agreed projects: BSTS, BVAR, dynamic linear/state-space models and Bayesian AR specifications using Stan or PyMC. Outputs can include posterior summaries, credible intervals and posterior predictive checks.
5. Which machine learning methods can you use for time series forecasts?
I can engineer lag, calendar and rolling features for SVR, Random Forest, XGBoost, LightGBM, CatBoost and Prophet. I compare ML models with statistical baselines using time-aware validation to avoid future-data leakage.
6. Can you develop LSTM, GRU, Transformer and other deep forecasts?
For data-rich, well-defined projects I can assess LSTM, GRU, TCN, N-BEATS, N-HiTS and Transformer/TFT approaches. Chronos or TimesFM and other pretrained models may be discussed as custom projects.
7. Can you analyze multiple series, exogenous variables and hybrid models?
Yes. Depending on scope: univariate/multivariate forecasting, external regressors, grouped series, intermittent demand, ARIMA-ML hybrids and forecast ensembles. The number of series and required horizon affect the quote.
8. How do you validate forecasting accuracy and quantify uncertainty?
I use chronological holdouts, rolling-origin backtesting and suitable MAE, RMSE, MASE, sMAPE or interval metrics. I check residuals, calibration and prediction intervals where appropriate; no model guarantees accuracy.
9. What software, packages and deliverables will I receive?
Tools include R/RStudio (forecast, fable, tsibble, rugarch), Python (statsmodels, sktime, scikit-learn, PyMC, PyTorch), Jupyter and Quarto. Deliverables may include scripts, fitted models, figures, reports and dashboards.
10. What should I send, and can I get a personalized custom offer?
Send your CSV/Excel data, date column, frequency, target, forecast horizon, predictors, methods and outputs. Basic is a paid assessment; advanced Bayesian, econometric, deep learning, multiple-series and deployment work is quoted separately.
1 reviews for this Gig
| (1) | ||
| (0) | ||
| (0) | ||
| (0) | ||
| (0) |
Rating Breakdown
- Seller communication level
- Quality of delivery
- Value of delivery
Sort By
A aykhanpolad177

Azerbaijan
Great teacher
$50-$100
Price
1 day
Duration
Helpful?
1 reviews for this Gig
| (1) | ||
| (0) | ||
| (0) | ||
| (0) | ||
| (0) |
Rating Breakdown
- Seller communication level
- Quality of delivery
- Value of delivery
Sort By
A aykhanpolad177

Azerbaijan
Great teacher
$50-$100
Price
1 day
Duration
Helpful?

