In today's financial markets, where most trades are performed in their entirety by electronic means and the largest fraction of them is completely automated, an opportunity has risen from analyzing this vast amount transactions. Since all transactions recorded great detail, investors can analyze generated data detect repeated patterns price movements. Being able to advance, allows take profitable positions or avoid anomalous events markets. work we proposed a deep learning methodology, based...

Financial time-series forecasting has long been a challenging problem because of the inherently noisy and stochastic nature market. In high-frequency trading, for trading purposes is even more task, since an automated inference system required to be both accurate fast. this paper, we propose neural network layer architecture that incorporates idea bilinear projection as well attention mechanism enables detect focus on crucial temporal information. The resulting highly interpretable, given...

Forecasting financial time-series has long been among the most challenging problems in market analysis. In order to recognize correct circumstances enter or exit markets investors usually employ statistical models (or even simple qualitative methods). However, inherently noisy and stochastic nature of severely limits forecasting accuracy used models. The introduction electronic trading availability large amounts data allow for developing novel machine learning techniques that address some...

Deep learning (DL) models can be used to tackle time series analysis tasks with great success. However, the performance of DL degenerate rapidly if data are not appropriately normalized. This issue is even more apparent when for financial forecasting tasks, where nonstationary and multimodal nature pose significant challenges severely affect models. In this brief, a simple, yet effective, neural layer that capable adaptively normalizing input series, while taking into account distribution...

Nowadays, with the availability of massive amount trade data collected, dynamics financial markets pose both a challenge and an opportunity for high frequency traders. In order to take advantage rapid, subtle movement assets in High Frequency Trading (HFT), automatic algorithm analyze detect patterns price change based on transaction records must be available. The multichannel, time-series representation naturally suggests tensor-based learning algorithms. this work, we investigate...

The existing literature provides evidence that limit order book data can be used to predict short-term price movements in stock markets. This paper proposes a new neural network architecture for predicting return jump arrivals one minute ahead equity markets with high-frequency data. architecture, based on Convolutional Long Short-Term Memory Attention, is introduced apply time series representation learning memory and focus the prediction attention most important features improve...

Abstract Managing the prediction of metrics in high‐frequency financial markets is a challenging task. An efficient way by monitoring dynamics limit order book to identify information edge. This paper describes first publicly available benchmark dataset for mid‐price prediction. We extracted normalized data representations time series five stocks from Nasdaq Nordic stock market period 10 consecutive days, leading ∼4,000,000 samples total. A day‐based anchored cross‐validation experimental...

Forecasting the movements of stock prices is one most challenging problems in financial markets analysis. In this paper, we use machine learning (ML) algorithms for prediction future price using limit order book data. Two different sets features are combined and evaluated: handcrafted based on raw data extracted by ML algorithms, resulting feature vectors with highly variant dimensionalities. Three classifiers evaluated combinations these two evaluation setups three scenarios. Even though...

Mid-price movement prediction based on the limit order book data is a challenging task due to complexity and dynamics of book. So far, there have been very limited attempts for extracting relevant features data. In this paper, we address problem by designing new set handcrafted performing an extensive experimental evaluation both liquid illiquid stocks. More specifically, present econometric that capture statistical properties underlying securities mid-price prediction. The consists...

Abstract Emergency department (ED) crowding is a global public health issue that has been repeatedly associated with increased mortality. Predicting future service demand would enable preventative measures aiming to eliminate along its detrimental effects. Recent findings in our ED indicate occupancy ratios exceeding 90% are 10-day In this paper, we aim predict these crisis periods using retrospective time series data such as weather, availability of hospital beds, calendar variables and...

Planning new product development ( NPD ) activities is becoming increasingly difficult, as contemporary businesses compete at the level of business ecosystems in addition to firm‐level product‐market competition. These are built around platforms interlinking suppliers, complementors, distributors, developers, etc. together. The competitiveness these relies on members utilizing shared platform for their own performance improvement, especially terms developing valuable offerings end users....

Time-series forecasting has various applications in a wide range of domains, e.g., stock markets using limit order book data. Limit data provide much richer information about the behavior stocks than its price alone, but also bear several challenges, such as dealing with multiple depths and processing very large amounts high dimensionality, velocity, variety. A well-known approach for efficiently handling high-dimensional is bag-of-features (BoF) model. However, BoF method was designed to...

Classification of time-series data is a challenging problem with many real-world applications, ranging from identifying medical conditions electroencephalography (EEG) measurements to forecasting the stock market. The well known Bag-of-Features (BoF) model was recently adapted towards representation. In this work, neural generalization BoF model, composed an RBF layer and accumulation layer, proposed as that receives features extracted gradually builds its method can be combined any other or...

Multilayer networks are attracting growing attention in many fields, including finance. In this paper, we develop a new tractable procedure for multilayer aggregation based on statistical validation, which apply to investor networks. Moreover, propose two other improvements their analysis: transaction bootstrapping and categorization. The can be used integrate security-wise time-wise information about trading networks, but it is not limited fact, different applications, such as gene,...

Relatively little is known about the empirical performance of infinite-activity Lévy jump models, especially with non-affine volatility dynamics. We use extensive data sets to study how Variance Gamma and Normal Inverse Gaussian (NIG) jumps affine dynamics improve goodness fit option pricing performance. With Markov Chain Monte Carlo, different model specifications are estimated using joint information S&P 500 index VIX. Our article provides clear evidence that a parsimonious NIG return...

Deep Learning models have become dominant in tackling financial time-series analysis problems, overturning conventional machine learning and statistical methods. Most often, a model trained for one market or security cannot be directly applied to another due differences inherent the conditions. In addition, as evolves over time, it is necessary update existing train new ones when data made available. This scenario, which most forecasting applications, naturally raises following research...

We identify temporal investor networks for Nokia stock by constructing from correlations between investor-specific net-volumes and analyze changes in the around dot-com bubble. conduct analysis separately households, non-financial institutions, financial institutions. Our results indicate that spanning tree measures households reflected boom crisis: maximum had clear upward tendency bull markets when bubble was building up, and, even more importantly, minimum pre-reacted burst of At same...

Recent studies using data on social media and stock markets have mainly focused predicting returns. Instead of price movements, we examine the relation between Facebook investors' decision making in with a unique transactions Nokia. We find that decisions to buy versus sell are associated especially for passive households also nonprofit organizations. At same time, it seems more sophisticated investors---financial insurance institutions---are behaving independently from activities.