Predicting the Future: Time Series and Data Science Part One: with applications in Python (Data Science and Engineering - A learning path) - Formato Tapa blanda
Disponible.
Precio:
Q345.00
Garantía:
Verifica las políticas de Tienda Mundial
Producto de Tienda Mundial
Ocultar detalles
Producto de importación, vendido por Amazon y transportado desde Miami. Cómpralo hoy y recibe en los próximos 6-14 días hábiles.
- This work covers the activities of Acquisition, Cleaning, Exploratory Data Analysis, Decomposition, Modelling and Forecasting of Time Series.It follows the 2021 curriculum of the Association for Computing Machinery for specialists in Data Sciences, with the aim of producing a manual that collects notions in a simplified form, facilitating a personal training path starting from specialized skills in Computer Science or Mathematics or Statistics. It has a bibliography with links to quality material but freely usable for your own training and contextual practical exercises.Since this text uses Python for application aspects, its installation and use are briefly described. In any case, this text should not be considered a Python manual. If by chance you made the wrong choice because you expected something different, you are free to return it but perhaps it is not the case to give a negative rating.The text first summarizes the standard CRISP DM working methodology used in this work and in Data Science projects. It then describes the fundamental concepts and terminology related to Time Series, data preprocessing and cleaning, handling missing and repeated values, managing extreme and outlier values, scaling and normalization.Exploratory Data Analysis (EDA) is explained in various possible scenarios, examining various diagnostic graphs and learning how to produce them in Python. The measures and metrics of location, dispersion, skewness, correlation, proximity, test and score, and similarity are then considered in theory and practice.Additive and multiplicative decomposition, detrending and deseasonalization with the various possible algorithms, stationarity and associated tests, transformations on time series and total and partial autocorrelations are then described.The modeling phase description begins with classic forecasting models (Naive, Moving Averages, and Smoothing), then it considers ARIMA and SARIMA models, explaining how they compare using the AIC, BIC, and MDL indexes. Advanced techniques consider ETS and State-Space models, associating them with Markov processes and the Kalman filter, including automatic AUTOARIMA and AUTOETS techniques. GAS, ARCH, and GARCH models are then explored for high-variability applications such as trading and finance.Dynamic regression models (DRM), in which a time series is influenced by exogenous variables, are explored in detail in their possible variations: DRM with ARIMA errors, Dynamic Harmonic Regression, Distributed Delay Regressive and Autoregressive and Transfer Function. Finally, machine learning Metas Prophet model is described.The exercises are written in Python using the Scikit-learn, Statsmodel, Sktime, Prophet, Pywavelet, and Statsforecast libraries. The text is accompanied by supporting materials, and Python examples and test data are available for download.
* Precio y unidades sujetos a revisión de Pacifiko
Información del producto
| PID | ZWQ4NzU0MW |
| Número de modelo | B0FQBWHX14 |
| Garantía | 7 días sujeto a las políticas de Tienda Mundial |
Garantía y Soporte
Para más información sobre Garantías en Pacifiko visitar la siguiente pagina: Garantías
Política de Igualación de Precios
¿Encontraste este producto más barato en Guatemala? Te igualamos el precio. Aplican condiciones.
Escribir por WhatsApp