Machine learning enhanced Hankel dynamic-mode decomposition.

While the acquisition of time series has become more straightforward, developing dynamical models from time series is still a challenging and evolving problem domain. Within the last several years, to address this problem, there has been a merging of machine learning tools with what is called the dynamic-mode decomposition (DMD). This general approach has been shown to be an especially promising avenue for accurate model development. Building on this prior body of work, we develop a deep learning DMD based method, which makes use of the fundamental insight of Takens' embedding theorem to build an adaptive learning scheme that better approximates higher dimensional and chaotic dynamics. We call this method the Deep Learning Hankel DMD. We likewise explore how our method learns mappings, which tend, after successful training, to significantly change the mutual information between dimensions in the dynamics. This appears to be a key feature in enhancing DMD overall, and it should help provide further insight into developing other deep learning methods for time series analysis and model generation.

Deep learning enhanced dynamic mode decomposition.

Koopman operator theory shows how nonlinear dynamical systems can be represented as an infinite-dimensional, linear operator acting on a Hilbert space of observables of the system. However, determining the relevant modes and eigenvalues of this infinite-dimensional operator can be difficult. The extended dynamic mode decomposition (EDMD) is one such method for generating approximations to Koopman spectra and modes, but the EDMD method faces its own set of challenges due to the need of user defined observables. To address this issue, we explore the use of autoencoder networks to simultaneously find optimal families of observables, which also generate both accurate embeddings of the flow into a space of observables and submersions of the observables back into flow coordinates. This network results in a global transformation of the flow and affords future state prediction via the EDMD and the decoder network. We call this method the deep learning dynamic mode decomposition (DLDMD). The method is tested on canonical nonlinear data sets and is shown to produce results that outperform a standard DMD approach and enable data-driven prediction where the standard DMD fails.

Phytotoxicity assays with hydroxyapatite nanoparticles lead the way to recover firing range soils.

Shooting activities is an important source of Pb in contaminated soils. Lead accumulates in superficial soil horizons because of its low mobility, favouring its uptake by plants and representing a high transference risk to the trophic chain. A combination of phytoremediation with nanoremediation techniques can be used to recover firing range soils and decrease the mobility, bioavailability and toxicity of Pb. This study examines in depth the changes in Pb behaviour in firing range soils by adding hydroxyapatite nanoparticles (HANPs). These nanoparticles (NPs) may immobilise Pb and improve the quality of these areas. The use of HANPs and the Pb effects were assessed in three different species (Sinapis alba L., Lactuca sativa L. and Festuca ovina L.), focusing on their germination and early growth, through phytotoxicity assays. Single extractions with CaCl2 (0.01 M) in soils treated with HANPs show that these NPs retained Pb and reduced highly its availability and mobility. HR-TEM and TOF-SIMS were used to determine the interactions between HANPs and Pb, as well as with soil components. According to TOF-SIMS and HR-TEM/EDS analysis, Pb was mainly retained by HANPs but also associated lightly to organic matter, Fe compounds and silicates. Phytotoxicity assays exposed that S. alba, L. sativa and F. ovina were able to germinate and develop in the firing range soils despite the high available Pb contents before adding HANPs. After adding HANPs, Pb retention increased, favouring the germination and the growth of roots in the three species. These results suggest that HANPs can be used to decrease the availability and the toxicity of Pb without negative effects in the species growth. Accordingly, the combination of phytoremediation and nanoremediation techniques can be a great tool to stabilise these soils, avoiding the Pb transfer to nearby areas and its entry in the trophic chain.

Using Ca3(PO4)2 nanoparticles to reduce metal mobility in shooting range soils.

Shooting activities are a very important source of contamination as they are commonly detected high concentrations of Pb in the soils from these facilities. Different remediation methods imply the immobilization of the pollutants by decreasing their mobility and availability and nanotechnology is a promising technique in this field. The effectiveness of calcium phosphate nanoparticles (CPNs) in the remediation of small-arms firing range and trap shooting range soils is evaluated in this work. The operationally defined extractable content of Pb, Cu and Zn is determined together with the interaction of the pollutants with the nanomaterials. Soil samples were treated with the CPNs and after the treatment the extractable contents of Cu, Pb and Zn decrease. To check the retention by the nanoparticles TOF-SIMS (Time of Flight Secondary Ion Mass Spectrometry) and HR-TEM-EDS (High Resolution Transmission Electron Microscopy with Energy Dispersive X-ray Spectroscopy) techniques were applied. The association of Pb and Cu to the CPNs was demonstrated by TOF-SIMS although it also indicated that not all the Pb and Cu contents are linked to the nanoparticles. By means of HR-TEM/EDS it was made out the filamentous shape and the size (50-150nm long and 20-40nm wide) of the CPNs together with their elemental composition (Ca, P and O). The CPNs were identified in treated soil samples together with signals of metals. The decrease on metal extractability detected is, in part, due to the association with CPNs but still more investigation is needed regarding mobility and availability of potentially hazardous elements in soils treated with nanoparticles.

An unusual case of breast cancer relapse after 30 years of disease-free survival.

A woman treated for breast cancer with mastectomy and radiotherapy in 1965 presented with a pleural mass and high CA 15.3 concentrations in a pleural effusion 30 years later and responded to endocrine therapy. The high CA 15.3 concentration was useful in diagnosis and treatment selection.

Rurality and prescription drug utilization among the elderly: an archival study.

Despite documentation that rural elderly have reduced access to both primary care and specialist physician services, there have been very few studies comparing rural and urban patterns of prescription drug use. This is unfortunate, because prescription drugs are the most commonly used type of health care by the elderly. This research merged claims data for a random sample of 18,641 enrolled elderly in the Pennsylvania Pharmaceutical Assistance Contract for the Elderly (PACE) for the years 1984 through 1988 with Medicare inpatient and outpatient health services records and with county-level demographic and health services resources data bases to test several models of factors associated with prescription drug use. The Human Resources Profile County Code from 1980 census data (HRPCC80) in the Area Resource File provided a very detailed (10 levels) definition of rurality. Consistent with our hypotheses based on preliminary studies, neither rurality designations nor county-level health care resource indices, nor interaction terms of health services resources with rurality were powerful predictors of prescription drug use. Use of health services (from Medicare data) and variables of longevity and continuity in the PACE program were consistently robust predictors of prescription drug use. Personal demographic characteristics were also strong predictors: white widowed women under age 85 with relatively higher incomes used more prescription drugs.

Empirical perspectives on services integration and aging: a review.

This article combines a review of social service literature with interviews of key informants on the issue of integrating services for the elderly into human services across the life span. Empirical findings concerning the relative benefits and drawbacks of age-integrated vs. specialized service organizational structures are emphasized. Surprisingly few empirical studies were identified. Instead of empirically based reports, the literature is largely comprised of polemics, which essentially restate long-standing positions on the preferred handling of integration. A persistent difficulty in discussions of federal and state service integration objectives has been the wide differences in definitions of "integration." The lack of data demonstrating that older persons benefit in integrated service systems, along with substantial progress in development of specialized services for the aged, suggest that aging advocates should be cautious in embracing age-integrated service integration proposals. Research within management information systems oriented toward client outcomes is necessary to gain valid knowledge of the merits of various integrated service models.

State pharmaceutical assistance for the elderly and disabled: planning issues for program and nonprogram states.

The Medicare Catastrophic Coverage Act (MCCA) would have mandated federal assistance for Medicare beneficiaries who have high annual prescription medication costs. High national expenditures for such drugs have encouraged the development of private and state insurance programs to help with these costs. Ten state pharmaceutical assistance programs (SPAPs), designed to help certain elderly, low income, or disabled people, exist for those ineligible for Medicaid or unable to purchase coverage privately. Coordination of state and federal benefits was a consideration for established programs, and programs being planned needed to determine the feasibility of integration of federal assistance. But the enactment and subsequent appeal of the Act affected both planning and policy implications for these SPAPs. All U.S. states and territories were surveyed before the bill's repeal to collect data on the effects of MCCA for those with prescription drug programs and those without. The repeal of the federal program places pressure on the nonprogram states to proceed, perhaps more cautiously, to initiate programs for their own residents, given increasing out-of-pocket and insurance costs, and no federal program.

Nutritional assessment of homebound rural elderly.

Approximately one quarter of Americans aged 65 and older live in rural areas. Homebound rural elderly are a hard to reach, often poor and frail population. Failure to include homebound rural elderly in nutritional studies may bias representativeness of samples of elderly persons. This paper discusses some challenges in recruiting homebound older persons to participate in nutritional studies and in establishing quality control procedures in a field setting.

Prescription drug coverage and medical indigence among the elderly.

When Medicare extends catastrophic coverage to outpatient prescription drugs in 1991, the program will benefit an estimated 15% to 25% of the aged population. For some of these individuals Medicare coverage will mean the difference between economic self-sufficiency and impoverishment due to high medical bills. This article provides empirical estimates of the relationship between third-party coverage of outpatient pharmaceuticals and the risk of medical indigence based on the experience in Pennsylvania following enactment of that state's Pharmaceutical Assistance Contract for the Elderly (PACE), a lottery-financed drug benefit program that currently enrolls nearly half a million residents age 65 and older. The study provides data on the distribution of billed charges for outpatient drugs by income class for PACE enrollees and then uses a time series analysis to estimate the impact of PACE implementation (July 1984) on monthly changes in Medicaid enrollments and expenditures by category of elderly Medicaid recipient over the period July 1981 through June 1987. Study findings are discussed in the context of the forthcoming Medicare drug benefit.

Use of a state-level pharmaceutical assistance program among urban and rural elders.

A number of state-level pharmaceutical assistance programs have been established as a result of the growing recognition of the role of pharmaceuticals in the long-term care of the elderly. However, existing research does not provide a coherent expectation for patterns of use by rural and urban elderly. The data for this analysis are drawn from a larger study of the Pennsylvania Pharmaceutical Assistance Contract for the Elderly (PACE). PACE provides prescription medicines for elderly who meet income requirements. The research project was designed to assess the characteristics of PACE program participants and non-participants on a wide range of issues. Chi-square analysis and regression models were used to assess the association between rural and urban residence and access to the PACE Program. The results indicate that rural/urban status of the elderly is not a significant predictor of the use of PACE. Other traditional variables (e.g., health self-rating and physician visits) did predict difference in the pattern of use.

State-level pharmaceutical assistance programs for the elderly: a national survey.

In order to put Pennsylvania's Pharmaceutical Assistance Contract for the Elderly (PACE) Program in a national context, a nationwide mail survey and telephone follow-up to each of the 58 State Unit Directors on Aging in the United States and its territories identified 10 programs. The results reported in this article are specific to the seven state-level pharmaceutical assistance programs which were in operation during the fiscal year 1984-85. In general, the programs varied on select program characteristics and on their efforts to address major policy issues. Data from the non-program states indicated support, legislative efforts, and a high interest in fiscal concerns. The findings reflect a lack of program uniformity and have implications for program development and implementation. Suggestions on how to identify the "optimum" or best combination of program and policy options are discussed.

On the social ecology of dependence and independence in elderly nursing home residents: a replication and extension.

A sample of 40 elderly residents was observed for 14 days on each of three consecutive occasions separated by 3 months. The observations focused on interactions involving the elderly resident and social partners. The results indicated a high level of stability across occasions in resident behaviors. The interactional and behavioral profiles observed were similar across length of institutionalization, sex, and health status of residents for the institutional context studied. Examination of interactional patterns for the different resident behaviors, however, revealed quite discrepant social contingencies. Following dependent self-care behaviors of residents, interactional responses by social partners represented a continuous supportive behavior contingency; following independent self-care behaviors, no social contingencies were observed. Similarly, obstructively engaged and nonengaged resident behaviors were not followed by any social contingencies; following constructively engaged behaviors of residents, an intermittent supportive contingency by social partners was found. The results are discussed in relation to learned helplessness and control.