---
title: Historical calibration of SVJD models with deep learning
url: https://www.ml-quant.com/papers/ssrn/4650097/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-09-26
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
identifier: SSRN 4650097
source_url: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4650097
featured: 2023-12-06
citations: 1
topic: ML & AI Methods
---


# Historical calibration of SVJD models with deep learning

The paper suggests using deep neural networks to calibrate parameters of Stochastic Volatility Jump Diffusion models, proving to be more accurate, robust, and faster than other methods.

- Source: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4650097
- Identifier: SSRN 4650097
- Released: 2023-12-01
- First featured: Quant Letter No. 28 (2023-12-06): https://www.ml-quant.com/issues/2023-12-06/
- Citations (Semantic Scholar): 1
- Published in: not yet
- Topic: ML & AI Methods

## Related

- [Edge Directionality Improves Learning on Heterophilic Graphs](https://www.ml-quant.com/papers/arxiv/2305.10498/): The study presents Directed Graph Neural Network (Dir-GNN), a new deep learning framework for directed graphs that surpasses traditional models in heterophilic benchmarks.
- [Modular Duality in Deep Learning](https://www.ml-quant.com/papers/arxiv/2410.21265/): The article presents a new theory of modular dualization for general neural networks, providing a theoretical basis for fast and scalable training algorithms, potentially leading to a new generation of optimizers for neural architectures.
- [Idempotent Generative Network](https://www.ml-quant.com/papers/arxiv/2311.01462/): A new generative modeling method is suggested, using an idempotent neural network to project any input into a target data distribution.
- [Can overfitted deep neural networks in adversarial training generalize? - An approximation viewpoint](https://www.ml-quant.com/papers/arxiv/2401.13624/): The study offers a theoretical insight into the robust overfitting issue in adversarial training on over-parameterized deep neural networks (DNNs), showing that overfitting can be avoided and a robust generalization gap is unavoidable, with the model capacity requirement depending on the target function's smoothness.
- [Scenario generation for market risk models using generative neural networks](https://www.ml-quant.com/papers/arxiv/2109.10072/): The study extends the use of generative adversarial networks (GANs) to a full internal market risk model, suggesting that GAN-based models can be a data-driven alternative for market risk modeling.
- [Evaluating Adversarial Robustness: A Comparison Of FGSM, Carlini-Wagner Attacks, And The Role of Distillation as Defense Mechanism](https://www.ml-quant.com/papers/arxiv/2404.04245/): The study investigates adversarial attacks on Deep Neural Networks for image classification, highlighting the Fast Gradient Sign Method and the Carlini-Wagner approach, and suggests defensive distillation as a defense, effective against FGSM but vulnerable to CW attacks.