Study on the Ground-Based FTS Measurements at Beijing, China and the Colocation Sensitivity of Satellite Data

The Fourier Transform Spectrometer (FTS) at the Beijing Satellite Meteorological Ground Station observed XCO2 (the dry carbon dioxide column) from 2 March 2016 to 4 December 2018. The validation results of ground-based XCO2, as well as GOSAT, OCO-2, and TanSat XCO2, show that the best temporal matching setting for ground-based XCO2 and satellite XCO2 is ±1 h, and the best spatial matching setting for GOSAT is 0.5° × 0.5°. Consistent with OCO-2, the best spatial matching setting of TanSat is 5° × 5° or 6° × 6°. Among GOSAT, OCO-2, and TanSat, the satellite observation validation characteristics near 5° × 5° from the ground-based station are obviously different from other spatial matching grids, which may be due to the different observation characteristics of satellites near 5° × 5°. To study the influence of local CO2 sources on the characteristics of satellite observation validation, we classified the daily XCO2 observation sequence into concentrated, dispersive, increasing, and decreasing types, respectively, and then validated the satellite observations. The results showed that the concentrated and decreasing sub-datasets have better validation performance. Our results suggest that it is best to use concentrated and decreasing sub-datasets when using the Beijing Satellite Meteorological Ground Station XCO2 for satellite validation. The temporal matching setting should be ±1 h, and the spatial matching setting should consider the satellites observation characteristics of 5° × 5° distance from the ground-based station.

Genetic Algorithm Approach to Find the Estimated Value of HMM parametersfor NS5 Methyltransferase Protein

Dengue is the pandemic disease caused by Dengue virus (DENV), a mosquito-borne flavivirus. In recent years dengue has emerged as a foremost cause of severe illness and deaths in developing countries.About 400 million dengue infections occur worldwide each year.In general, dengue infections create only mild illness but infrequently expand into a lethal illness termed as severe dengue for which no specific treatment. The machine learning approach plays a significant role in bioinformatics and other fields of computer science.It exploitsapproaches like Hidden Markov Model (HMM), Genetic Algorithm (GA), Artificial Neural Network (ANN), and Support Vector Machine (SVM).The GA is a randomized search algorithm for solving the problem based on natural selection phenomena.Many machine learning techniques are based on HMM have been positively applied. In this work, We firstly used HMM parameters on the biological sequence,and after that, we catch the probability of the observation sequence of a mutated gene sequence. This study comparesboth methods, G.A. and HMM, to get the highest estimated value of the observation sequence. In this paper, we also discuss the applications ofGA in the bioinformatics field. In a further study, we will apply the other machine learning approaches to find the best result of protein studies.

Prediction of Chaotic Time Series Based on BEN-AGA Model

Aiming at the prediction problem of chaotic time series, this paper proposes a brain emotional network combined with an adaptive genetic algorithm (BEN-AGA) model to predict chaotic time series. First, we improve the emotional brain learning (BEL) model using the activation function to change the two linear structures the amygdala and the orbitofrontal cortex into the nonlinear structure, and then we establish the brain emotional network (BEN) model. The brain emotional network model has stronger nonlinear calculation ability and generalization ability. Next, we use the adaptive genetic algorithm to optimize the parameters of the brain emotional network model. The weights to be optimized in the model are coded as chromosomes. We design the dynamic crossover probability and mutation probability to control the crossover process and the mutation process, and the optimal parameters are selected through the fitness function to evaluate the chromosome. In this way, we increase the approximation capability of the model and increase the calculation speed of the model. Finally, we reconstruct the phase space of the observation sequence based on the short-term predictability of the chaotic time series; then we establish a brain emotional network model and optimize its parameters with an adaptive genetic algorithm and perform a single-step prediction on the optimized model to obtain the prediction error. The model proposed in this paper is applied to the prediction of Rossler chaotic time series and sunspot chaotic time series. The experimental results verify the effectiveness of the BEN-AGA model and show that this model has higher prediction accuracy and more stability than other methods.

Planning of refracted starlight observation based on observability analysis

In traditional observation schemes of stellar refraction navigation, the accuracy was limited due to unreasonable observation directions. In order to ameliorate this situation, a method of refracted starlight observation based on observability analysis is proposed. The function of this method is optimally generating an observation attitude sequence according to standard trajectories of spacecraft so that the selection of a refracted starlight observation sequence can be realized. Specifically, the improvement of Fisher information matrix calculation enables this method to be qualified for the navigation problem with unsteady measurement quantities as well as the non-fully observability which is defined as the capability of estimating the system state through measurements in finite time. Here, we construct a quantitative relationship between refracted starlight measurements and system observability by means of Fisher information index ( FII). Next, the observation scheme is retrieved by searching the maximum value of the optimized variable, which includes the ( FII). Finally, we resort to the extended Kalman filter to accomplish typical trajectory navigation simulations of the observation scheme. The results indicate that our method brings more accuracy than traditional ones in estimation of position and velocity of the optimal observation scheme.

The Convergence Rate of High-Dimensional Sample Quantiles for φ-Mixing Observation Sequences

In this paper, we obtain the convergence rate for the high-dimensional sample quantiles with the φ-mixing dependent sequence. The resulting convergence rate is shown to be faster than that obtained by the Hoeffding-type inequalities. Moreover, the convergence rate of the high-dimensional sample quantiles for the observation sequence taking discrete values is also provided.

Observation Time Effects in Reinforcement Learning on Contracts for Difference

In this paper, we present a study on Reinforcement Learning optimization models for automatic trading, in which we focus on the effects of varying the observation time. Our Reinforcement Learning agents feature a Convolutional Neural Network (CNN) together with Long Short-Term Memory (LSTM) and act on the basis of different observation time spans. Each agent tries to maximize trading profit by buying or selling one of a number of contracts in a simulated market environment for Contracts for Difference (CfD), considering correlations between individual assets by architecture. To decide which action to take on a specific contract, an agent develops a policy which relies on an observation of the whole market for a certain period of time. We investigate whether or not there exists an optimal observation sequence length, and conclude that such a value depends on market dynamics.

Learning in Volatile Environments with the Bayes Factor Surprise

Surprise-based learning allows agents to rapidly adapt to nonstationary stochastic environments characterized by sudden changes. We show that exact Bayesian inference in a hierarchical model gives rise to a surprise-modulated trade-off between forgetting old observations and integrating them with the new ones. The modulation depends on a probability ratio, which we call the Bayes factor surprise, that tests the prior belief against the current belief. We demonstrate that in several existing approximate algorithms, the Bayes Factor Surprise modulates the rate of adaptation to new observations. We derive three novel surprise-based algorithms, one in the family of particle filters, one in the family of variational learning, and one in the family of message passing, that have constant scaling in observation sequence length and particularly simple update dynamics for any distribution in the exponential family. Empirical results show that these surprise-based algorithms estimate parameters better than alternative approximate approaches and reach levels of performance comparable to computationally more expensive algorithms. The Bayes Factor Surprise is related to but different from the Shannon Surprise. In two hypothetical experiments, we make testable predictions for physiological indicators that dissociate the Bayes factor surprise from the Shannon Surprise. The theoretical insight of casting various approaches as surprise-based learning, as well as the proposed online algorithms, may be applied to the analysis of animal and human behavior and to reinforcement learning in nonstationary environments.

Flying through a dawn storm: a multi-instrument study of the traversal of a dawn storm by Juno on February 7th 2018

<p class="western" align="left">On February 7th 2018, during Juno’s 11th perijove observation sequence, Juno’s ultraviolet spectrograph (Juno-UVS) unveiled the development of a dawn storm in Jupiter's aurorae. These auroral events consist of spectacular brightenings of the midnight to dawn sector of the main emissions at Jupiter. At the end of the sequence, Juno crossed the magnetic field lines connected to this dawn storm, unraveling some of the processes giving rise to these spectacular events. <br />All in situ instruments detected a sharp transition as the spacecraft entered the dawn storm at an altitude of approximately 5RJ in the southern hemisphere. The particle fluxes detected by the JADE and JEDI instruments, including electrons and ions, increased dramatically. A strong flux of penetrating radiation was also detected by the UVS instrument. The Alfvén waves spectrograms derived from the MAG instrument also show a clear transition between a quiet and an extremely active regime as the spacecraft entered the dawn storm. Furthermore, the orientation of the magnetic field showed a very strong perturbation, associated with intense currents. And, finally, intense bKOM emissions were also observed during this time interval. Combined with the remote sensing observations of the aurora, these datasets strongly suggest that Juno witnessed a strong magnetospheric reconfiguration that started in the magneto-tail and then evolved toward dawn as the planet rotated.</p>