Source author record

Sang Il Lee

Sang Il Lee appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

q-fin.CP cond-mat.stat-mech Machine Learning q-fin.PM

Catalog footprint

What is connected

3works

4topics

1close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2020arXiv

Deeply Equal-Weighted Subset Portfolios

The high sensitivity of optimized portfolios to estimation errors has prevented their practical application. To mitigate this sensitivity, we propose a new portfolio model called a Deeply Equal-Weighted Subset Portfolio (DEWSP). DEWSP is a subset of top-N ranked assets in an asset universe, the members of which are selected based on the predicted returns from deep learning algorithms and are equally weighted. Herein, we evaluate the performance of DEWSPs of different sizes N in comparison with the performance of other types of portfolios such as optimized portfolios and historically equal-weighed subset portfolios (HEWSPs), which are subsets of top-N ranked assets based on the historical mean returns. We found the following advantages of DEWSPs: First, DEWSPs provides an improvement rate of 0.24% to 5.15% in terms of monthly Sharpe ratio compared to the benchmark, HEWSPs. In addition, DEWSPs are built using a purely data-driven approach rather than relying on the efforts of experts. DEWSPs can also target the relative risk and return to the baseline of the EWP of an asset universe by adjusting the size N. Finally, the DEWSP allocation mechanism is transparent and intuitive. These advantages make DEWSP competitive in practice.

preprint2020arXiv

Hyperparameter Optimization for Forecasting Stock Returns

In recent years, hyperparameter optimization (HPO) has become an increasingly important issue in the field of machine learning for the development of more accurate forecasting models. In this study, we explore the potential of HPO in modeling stock returns using a deep neural network (DNN). The potential of this approach was evaluated using technical indicators and fundamentals examined based on the effect the regularization of dropouts and batch normalization for all input data. We found that the model using technical indicators and dropout regularization significantly outperforms three other models, showing a positive predictability of 0.53% in-sample and 1.11% out-of-sample, thereby indicating the possibility of beating the historical average. We also demonstrate the stability of the model in terms of the changes in its feature importance over time.

preprint2010arXiv

Two-dimensional melting in simple atomic systems: continuous vs. discontinous melting

We investigate the characteristics of two dimensional melting in simple atomic systems via isobaric-isothermal ($NPT$) and isochoric-isothermal ($NVT$) molecular dynamics simulations with special focus on the effect of the range of the potential on the melting. We find that the system with interatomic potential of longer range clearly exhibits a region (in the $PT$ plane) of (thermodynamically) stable hexatic phase. On the other hand, the one with shorter range potential exhibits a first-order melting transition both in $NPT$ and $NVT$ ensembles. Melting of the system with intermediate range potential shows a hexatic-like feature near the melting transition in $NVT$ ensemble, but it undergoes an unstable hexatic-like phase during melting process in $NPT$ ensemble, which implies existence of a weakly first order transition. The overall features represent a crossover from a continuous melting transition in the cases of longer-ranged potential to a discontinuous (first order) one in the systems with shorter and intermediate ranged potential. We also calculate the Binder cumulants as well as the susceptibility of the bond-orientational order parameter.