deepregression: a Flexible Neural Network Framework for Semi-Structured Deep Distributional Regression

Rügamer, David; Kolb, Chris; Fritz, Cornelius; Pfisterer, Florian; Kopper, Philipp; Bischl, Bernd; Ruolin, Shen,; Bukas, Christina; Sousa, Lisa Barros de Andrade e; Thalmeier, Dominik; Baumann, Philipp; Kook, Lucas; Klein, Nadja; Müller, Christian L.

doi:10.48550/arxiv.2104.02705

Cited by 6 publications

(7 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The additive predictor in our approach allows for straightforward interpretability and to recover the PAM(M) when no additional deep predictors are necessary. Our method can be fit using existing software solutions (e.g., deepregression [34]).…”

Section: Discussionmentioning

confidence: 99%

DeepPAMM: Deep Piecewise Exponential Additive Mixed Models for Complex Hazard Structures in Survival Analysis

Kopper¹,

Wiegrebe²,

Bischl³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Survival analysis (SA) is an active field of research that is concerned with time-to-event outcomes and is prevalent in many domains, particularly biomedical applications. Despite its importance, SA remains challenging due to smallscale data sets and complex outcome distributions, concealed by truncation and censoring processes. The piecewise exponential additive mixed model (PAMM) is a model class addressing many of these challenges, yet PAMMs are not applicable in high-dimensional feature settings or in the case of unstructured or multimodal data. We unify existing approaches by proposing DeepPAMM, a versatile deep learning framework that is well-founded from a statistical point of view, yet with enough flexibility for modeling complex hazard structures. We illustrate that DeepPAMM is competitive with other machine learning approaches with respect to predictive performance while maintaining interpretability through benchmark experiments and an extended case study.

show abstract

Section: Discussionmentioning

confidence: 99%

DeepPAMM: Deep Piecewise Exponential Additive Mixed Models for Complex Hazard Structures in Survival Analysis

Kopper¹,

Wiegrebe²,

Bischl³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Here, w is the weight of input traffic x 1 to x d , and b is known as bias or offset. Weight determines the influence of features in the model [36], [37] and [38].…”

Section: F Regression With Deep Neural Networkmentioning

confidence: 99%

Deep Learning Based Fusion Model for Multivariate LTE Traffic Forecasting and Optimized Radio Parameter Estimation

et al. 2023

View full text Add to dashboard Cite

With the evaluation of cellular network internet data traffic, forecasting and understanding traffic patterns become the critical objectives for managing the network-designed Quality of Service (QoS) benchmark. For this purpose, cellular network planners often use different methodologies for predicting data traffic. However, traditional traffic forecasting approaches are erroneous. As well as most of the time, traditional traffic forecasts are high-level or a generously large regional cluster level. Also, eNodeB-level utilization with concerning traffic forecasting is not readily available. As a result, user experience degradation or unnecessary network expansion is triggered based on the traditional method. This research deals with extensive 6.2 million real network time series Long-Term Evolution (LTE) data traffic and other associate parameters, including eNodeB-wise Physical Resource Block (PRB) utilization, which focuses on building a traffic forecasting model with the help of multivariate feature inputs and deep learning algorithms. A stateof-the-art Deep Learning algorithm-based fusion model is proposed. The combination of different deep learning algorithms, namely Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM) and Gated Recurrent Unit (GRU), enables traffic forecasting at a granular eNodeB-level and also provides eNodeB-wise forecasted PRB utilization. In this research R 2 score value for the proposed fusion model is 0.8034, which outperforms traditional models. Apart from the PRB utilization, QoS threshold was devised as 70% from a real network experience to trigger soft parameter tuning decisions. Based on the forecasted PRB utilization, this research proposed a unique algorithm that estimates eNodeB-level soft capacity parameter optimization for a short-term step-up solution or long-term network expansion to ensure a guaranteed QoS benchmark.

show abstract

“…Selain itu pada bidang kesehatan, model regresi juga terus menjadi kebutuhan utama dalam kasus prediksi suatu penyakit [8]. Didorong oleh kebutuhan untuk memecahkan masalah regresi nonlinier dan multidimensi di dunia industri, peneliti di seluruh dunia untuk berusaha untuk menggunakan metode penyelesaian data multidimensi yang non-linier dengan model machine learning yang lebih kuat, seperti jaringan syaraf tiruan [9], deep regression [10], [11], random linear target combination [12], variational autoencoder regression [13], support vector regression [14], dan lainnya. Tingginya kebutuhan untuk melakukan komputasi yang lebih kompleks, kemampuan komputer dalam memproses data juga harus ikut meningkat, terutama dalam permasalahan machine learning [15].…”

Section: Pendahuluanunclassified

Kombinasi Linier Target Data Untuk Regresi Multitarget Menggunakan Principal Component Analysis

Santosa

2023

j. teknologi terpadu

View full text Add to dashboard Cite

Regresi linier adalah metode untuk memprediksi sebuah nilai (variabel dependen) berdasarkan beberapa input (variabel independen). Permasalahan pada regresi linier adalah beberapa data tidak termasuk ke dalam kategori linier. Sebuah metode bernama RLC diciptakan untuk menemukan korelasi antara data output dengan cara memproyeksikan data ke dalam dimensi yang lebih tinggi. Sayangnya, metode RLC tidak dapat diinvers transformasinya. Selain itu, dengan memproyeksikan data ke dimensi yang lebih tinggi akan menambah kompleksitas dari algoritma pembelajaran. Oleh karena itu, PCA akan digunakan untuk memecahkan masalah ini dengan cara memproyeksikan data ke dimensi yang lebih rendah sembari mempertahankan kemampuan untuk melakukan invers proyeksi. Penelitian ini diimplementasikan dengan bantuan library scikit-learn untuk membuat model regresi dan transformasi data dengan menggunakan bahasa pemrograman Python. Hasilnya, untuk 12 dataset, metode augmentasi PCA mampu mendapatkan nilai error yang lebih rendah dalam 7 dataset dibandingkan dengan RLC dengan rata-rata nilai error 0.3270 untuk metode augmentasi PCA dan 0.4003 untuk metode augmentasi RLC.

show abstract

deepregression: a Flexible Neural Network Framework for Semi-Structured Deep Distributional Regression

Cited by 6 publications

References 2 publications

DeepPAMM: Deep Piecewise Exponential Additive Mixed Models for Complex Hazard Structures in Survival Analysis

DeepPAMM: Deep Piecewise Exponential Additive Mixed Models for Complex Hazard Structures in Survival Analysis

Deep Learning Based Fusion Model for Multivariate LTE Traffic Forecasting and Optimized Radio Parameter Estimation

Kombinasi Linier Target Data Untuk Regresi Multitarget Menggunakan Principal Component Analysis

Contact Info

Product

Resources

About