Data-driven Covariate Selection for Confounding Adjustment by Focusing on the Stability of the Effect Estimator

Valid inference of cause-and-effect relations in observational studies necessitates adjusting for common causes of the focal predictor (i.e., treatment) and the outcome. When such common causes, henceforth termed confounders, remain unadjusted for, they generate spurious correlations that lead to biased causal effect estimates. But routine adjustment for all available covariates, when only a subset are truly confounders, is known to yield potentially inefficient and unstable estimators. In this article, we introduce a data-driven confounder selection strategy that focuses on stable estimation of the treatment effect. The approach exploits the causal knowledge that after adjusting for confounders to eliminate all confounding biases, adding any remaining non-confounding covariates associated with only treatment or outcome, but not both, should not systematically change the effect estimator. The strategy proceeds in two steps. First, we prioritize covariates for adjustment by probing how strongly each covariate is associated with treatment and outcome. Next, we gauge the stability of the effect estimator by evaluating its trajectory adjusting for different covariate subsets. The smallest subset that yields a stable effect estimate is then selected. Thus, the strategy offers direct insight into the (in)sensitivity of the effect estimator to the chosen covariates for adjustment. The ability to correctly select confounders and yield valid causal inference following data-driven covariate selection is evaluated empirically using extensive simulation studies. Furthermore, we compare the proposed method empirically with routine variable selection methods. Finally, we demonstrate the procedure using two publicly available real-world datasets.

Data-driven Covariate Selection for Confounding Adjustment by Focusing on the Stability of the Effect Estimator

Valid inference of cause-and-effect relations in observational studies necessitates adjusting for common causes of the focal predictor (i.e., treatment) and the outcome. When such common causes, henceforth termed confounders, remain unadjusted for, they generate spurious correlations that lead to biased causal effect estimates. But routine adjustment for all available covariates, when only a subset are truly confounders, is known to yield potentially inefficient and unstable estimators. In this article, we introduce a data-driven confounder selection strategy that focuses on stable estimation of the treatment effect. The approach exploits the causal knowledge that after adjusting for confounders to eliminate all confounding biases, adding any remaining non-confounding covariates associated with only treatment or outcome, but not both, should not systematically change the effect estimator. The strategy proceeds in two steps. First, we prioritize covariates for adjustment by probing how strongly each covariate is associated with treatment and outcome. Next, we gauge the stability of the effect estimator by evaluating its trajectory adjusting for different covariate subsets. The smallest subset that yields a stable effect estimate is then selected. Thus, the strategy offers direct insight into the (in)sensitivity of the effect estimator to the chosen covariates for adjustment. The ability to correctly select confounders and yield valid causal inference following data-driven covariate selection is evaluated empirically using extensive simulation studies. Furthermore, we compare the proposed method empirically with routine variable selection methods. Finally, we demonstrate the procedure using two publicly available real-world datasets.

An audio enhancement system to improve intelligibility for social-awareness in HRI

Improving the ability to interact through voice with a robot is still a challenge especially in real environments where multiple speakers coexist. This work has evaluated a proposal based on improving the intelligibility of the voice information that feeds an existing ASR service in the network and in conditions similar to those that could occur in a care centre for the elderly. The results indicate the feasibility and improvement of a proposal based on the use of an embedded microphone array and the use of a simple beamforming and masking technique. The system has been evaluated with 12 people and results obtained for time responsiveness indicate that the system would allow natural interaction with voice. It is shown to be necessary to incorporate a system to properly employ the masking algorithm, through the intelligent and stable estimation of the interfering signals. In addition, this approach allows to fix as sources of interest other speakers not located in the vicinity of the robot.

Inflationary Effect on Nigeria Economic Growth: The ARDL Bound Test Approach

The expectation of any country is to experience a high output but in the presence of increasing inflation such expectation becomes blurring because high inflation is a sign of a low working economic system. In this research the impact of inflation rate (InfR) on Nigeria economic growth (EcoG) is studied for the period of 1986 to 2018 using an Autoregressive Distributed Lag (ARDL) Bounds test approach to determine the co-integration existence between InfR and EcoG and determine the long run effect through the approach of Error Correction Model (ECM). The results obtained showed that an ARDL (2, 2) model was the best fitted model for the sampled data based on the smallest Akaike’s Information Criterion (AIC) value obtained. Also, it was found that InfR significantly impacted on Nigeria EcoG negatively on the long and short run dynamics with a stable estimation as portrayed by the CUSUM square chart.

Analysis of Occlusion Effects for Map-Based Self-Localization in Urban Areas

A high-definition (HD) map provides structural information for map-based self-localization, enabling stable estimation in real environments. In urban areas, there are many obstacles, such as buses, that occlude sensor observations, resulting in self-localization errors. However, most of the existing HD map-based self-localization evaluations do not consider sudden significant errors due to obstacles. Instead, they evaluate this in terms of average error over estimated trajectories in an environment with few occlusions. This study evaluated the effects of self-localization estimation on occlusion with synthetically generated obstacles in a real environment. Various patterns of synthetic occlusion enabled the analyses of the effects of self-localization error from various angles. Our experiments showed various characteristics that locations susceptible to obstacles have. For example, we found that occlusion in intersections tends to increase self-localization errors. In addition, we analyzed the geometrical structures of a surrounding environment in high-level error cases and low-level error cases with occlusions. As a result, we suggested the concept that the real environment should have to achieve robust self-localization under occlusion conditions.

COMPARATIVE ANALYSIS OF SPECTRA OF THE 2000-YEAR NORTHERN HEMISPHERE AVERAGE SURFACE AIR TEMPERATURE RECONSTRUCTIONS

The spectral composition of temporal variations in the Northern Hemisphere mean surface air temperature is estimated and compared in 2000-year paleoclimatic reconstructions. Continuous wavelet transforms of these reconstructions are used for the stable estimation of energy spectra. It is found that low-frequency parts of the spectra (the periods of temperature variations of more than 100 years) based on such high-resolution paleoclimatic indicators as tree rings, corals, etc., are similar to the spectrum of white noise, that is never observed in nature. This seems unrealistic. The famous reconstruction called “Hockey Stick” is among such unrealistic reconstructions. Reconstructions based not only on high-resolution but also on low-resolution indicators seem to be more realistic, since the low-frequency parts of their spectra have the pattern of red noise. They include the “Boomerang” reconstruction showing that some warm periods close to the present-day one were observed in the past.

Using a New Circular Prediction Algorithm to Design an IMM Filter for Low Update Rate Radar System

For radar systems with low update rates; such as track-while-scan (TWS) systems using rotating phased array antennas; reducing the prediction error is a very important issue. A good interacting multiple models (IMM) hybrid filter combined with circular and linear filters that are defined in relation to three measurements has been proposed in the literature. However; the algorithm requires three previous measurements; and too much prior information will result in a reduced ability to predict the future position of a highly maneuvering target. A new circular prediction algorithm for maneuvering target tracking is proposed as a non-linear prediction filter in this paper. Based on this new predictor; we also proposed a new type of IMM filter that has good estimation performance for high maneuvering targets. The proposed hybrid filter is entirely defined in relation to two measurements in a three-dimensional space to obtain a better maneuver following capability than the three measurements hybrid filter. Two target profiles are included for a comparison of the performance of our proposed scheme with that of the conventional circular; linear and IMM filters. The simulation results show that under low update rates; the proposed filter has a faster and more stable estimation response than other filters

An innovative analytic redundancy approach to air data sensor fault detection

ABSTRACTThis article presents a potential analytic redundancy approach to detect faults in the air data sensor of an aircraft. In modern aircraft, fault detection of air data sensors is performed using a complex voting mechanism, which requires the availability of redundant air data sensor in all situations. However, to continuously monitor operation and performance of these sensors, the analytic redundancy-based air data estimation and fault detection is highly preferred than estimation with air data probe measurements. The proposed algorithm uses the kinematics of aircraft to estimate air data and detect air data sensor fault. In this paper, a simple mathematical model is developed, which does not consider the forces and moments acting on aircraft and uses measurements only from the Inertial Measurement Unit (IMU) and Navigation System Data (NSD). In order to implement this approach, the Iterated Optimal Extended Kalman Filter (IOEKF) is developed to estimate air data, which provides an accurate and stable estimation. With the estimated states, the physical air data sensor measurements are compared and the residual is calculated to track each sensor performance and to detect the occurrence of a fault. The key advantage of this approach is that it does not require complex dynamic equations and is free from system uncertainties. The proposed algorithm is simulated in MATLAB software using flight simulator flight data and validated using the real-time flight data of Cessna Citation II transport aircraft.