A domain-agnostic approach for characterization of lifelong learning systems

Despite the advancement of machine learning techniques in recent years, state-of-the-art systems lack robustness to "real world" events, where the input distributions and tasks encountered by the deployed systems will not be limited to the original training context, and systems will instead need to adapt to novel distributions and tasks while deployed. This critical gap may be addressed through the development of "Lifelong Learning" systems that are capable of (1) Continuous Learning, (2) Transfer and Adaptation, and (3) Scalability. Unfortunately, efforts to improve these capabilities are typically treated as distinct areas of research that are assessed independently, without regard to the impact of each separate capability on other aspects of the system. We instead propose a holistic approach, using a suite of metrics and an evaluation framework to assess Lifelong Learning in a principled way that is agnostic to specific domains or system techniques. Through five case studies, we show that this suite of metrics can inform the development of varied and complex Lifelong Learning systems. We highlight how the proposed suite of metrics quantifies performance trade-offs present during Lifelong Learning system development - both the widely discussed Stability-Plasticity dilemma and the newly proposed relationship between Sample Efficient and Robust Learning. Further, we make recommendations for the formulation and use of metrics to guide the continuing development of Lifelong Learning systems and assess their progress in the future.

Sample-Based Tree Search with Fixed and Adaptive State Abstractions

Sample-based tree search (SBTS) is an approach to solving Markov decision problems based on constructing a lookahead search tree using random samples from a generative model of the MDP. It encompasses Monte Carlo tree search (MCTS) algorithms like UCT as well as algorithms such as sparse sampling. SBTS is well-suited to solving MDPs with large state spaces due to the relative insensitivity of SBTS algorithms to the size of the state space. The limiting factor in the performance of SBTS tends to be the exponential dependence of sample complexity on the depth of the search tree. The number of samples required to build a search tree is O((|A|B)^d), where |A| is the number of available actions, B is the number of possible random outcomes of taking an action, and d is the depth of the tree. State abstraction can be used to reduce B by aggregating random outcomes together into abstract states. Recent work has shown that abstract tree search often performs substantially better than tree search conducted in the ground state space. This paper presents a theoretical and empirical evaluation of tree search with both fixed and adaptive state abstractions. We derive a bound on regret due to state abstraction in tree search that decomposes abstraction error into three components arising from properties of the abstraction and the search algorithm. We describe versions of popular SBTS algorithms that use fixed state abstractions, and we introduce the Progressive Abstraction Refinement in Sparse Sampling (PARSS) algorithm, which adapts its abstraction during search. We evaluate PARSS as well as sparse sampling with fixed abstractions on 12 experimental problems, and find that PARSS outperforms search with a fixed abstraction and that search with even highly inaccurate fixed abstractions outperforms search without abstraction. These results establish progressive abstraction refinement as a promising basis for new tree search algorithms, and we propose directions for future work within the progressive refinement framework.

Susceptibility of cat fleas (Siphonaptera: Pulicidae) to fipronil and imidacloprid using adult and larval bioassays

The monitoring of the susceptibility offleas to insecticides has typically been conducted by exposing adults on treated surfaces. Other methods such as topical applications of insecticides to adults and larval bioassays on treated rearing media have been developed. Unfortunately, baseline responses of susceptible strains of cat flea, Ctenocephalides felis (Bouchè), except for imidacloprid, have not been determined for all on-animal therapies and new classes of chemistry now being used. However, the relationship between adult and larval bioassays of fleas has not been previously investigated. The adult and larval bioassays of fipronil and imidacloprid were compared for both field-collected isolates and laboratory strains. Adult topical bioassays of fipronil and imidacloprid to laboratory strains and field-collected isolates demonstrated that LD50s of fipronil and imidacloprid ranged from 0.11 to 0.40 nanograms per flea and 0.02 to 0.18 nanograms per flea, respectively. Resistance ratios for fipronil and imidacloprid ranged from 0.11 to 2.21. Based on the larval bioassay published for imidacloprid, a larval bioassay was established for fipronil and reported in this article. The ranges of the LC50s of fipronil and imidacloprid in the larval rearing media were 0.07-0.16 and 0.11-0.21 ppm, respectively. Resistance ratios for adult and larval bioassays ranged from 0.11 to 2.2 and 0.58 to 1.75, respectively. Both adult and larval bioassays provided similar patterns for fipronil and imidacloprid. Although the adult bioassays permitted a more precise dosage applied, the larval bioassays allowed for testing isolates without the need to maintain on synthetic or natural hosts.

Efficacy of imidacloprid (8.8% w/w) plus permethrin (44% w/w) spot-on topical solution against Amblyomma americanum infesting dogs using a natural tick exposure model

This study evaluated the efficacy of an imidacloprid 8.8% w/w + permethrin 44% w/w spot-on topical solution (K9 Advantix, Bayer Animal Health) against Amblyomma americanum using a natural field exposure model. Sixteen beagles were divided into two groups of eight dogs each. One group of dogs was treated with K9 Advantix and the other group served as untreated controls. On day -1 and at 3, 7, 14, 21 and 28 days after treatment, the dogs were walked for 80 minutes in an A. americanum-infested habitat at the Konza Prairie Biological Station in Northeastern Kansas. Postexposure tick counts (efficacy evaluations) were conducted on each dog at 3 and 48 hours after exposure. At 3 days after treatment, the efficacy of K9 Advantix within 3 hours of natural tick exposure was 88.0% and declined slowly during the study. The 48-hour postexposure efficacy remained above 93.5% throughout the study.

Evaluation of an imidacloprid (8.8% w/w)--permethrin (44.0% w/w) topical spot-on and a fipronil (9.8% w/w)--(S)-methoprene (8.8% w/w) topical spot-on to repel, prevent attachment, and kill adult Rhipicephalus sanguineus and Dermacentor variabilis ticks on dogs

This study evaluated the effectiveness of two topical spot-on formulations, imidacloprid (8.8% w/w)--permethrin (44.0% w/w) and fipronil (9.8% w/w)--(S)-methoprene (8.8% w/w), to repel, prevent the attachment of, and kill adult Rhipicephalus sanguineus and Dermacentor variabilis on dogs. Twelve purpose-bred beagles were distributed into three groups of four dogs each; one group served as untreated controls and each of the other two groups received one of the test products. Dogs were exposed to 25 adult ticks of each species for 10 minutes on posttreatment days 3, 7, 14, 21, and 28. Nonattached or repelled ticks were collected and evaluated for viability, and on-dog tick counts were conducted at 3, 24, and 48 hours after tick exposure. The imidacloprid-permethrin formulation provided significant repellency of R. sanguineus and D. variabilis for up to 3 and 4 weeks after treatment, respectively; and provided good overall control for R. sanguineus and D. variabilis during the study period. The fipronil--(S)-methoprene formulation provided good overall tick control during the study period.

Evaluation of an imidacloprid (8.8% w/w)--permethrin (44.0% w/w) topical spot-on and a fipronil (9.8% w/w)--(S)-methoprene (8.8% w/w) topical spot-on to repel, prevent attachment, and kill adult Ixodes scapularis and Amblyomma americanum ticks on dogs

This study evaluated the effectiveness of two topical spot-on formulations -- imidacloprid(8.8% w/w)-permethrin (44.0% w/w) and fipronil (9.8% w/w)-(S)-methoprene (8.8% w/w)--to repel, prevent the attachment of, and kill adult Ixodes scapularis and Amblyomma americanum on dogs. Twelve purpose-bred beagles were distributed into three groups of four dogs each; one group served as untreated controls, and each of the other two groups received one of the test products. Dogs were exposed to 25 adult ticks of each species for 10 minutes on posttreatment days 3, 7, 14, 21, and 28. Unattached or repelled ticks were collected and evaluated for viability, and on-dog tick counts were conducted at 3, 24, and 48 hours after tick exposure. The imidacloprid-permethrin formulation provided significant repellency against I. scapularis for up to 3 weeks after treatment, and both formulations provided good overall control of I. scapularis and A. americanum during the study period.

Preventing substance abuse. An interview paradigm

In summary, substance use and abuse is a major health problem in the United States affecting adolescents of both sexes and of every socioeconomic level. Family physicians have numerous opportunities to improve the health of future generations by using specific skills and expertise to prevent, identify, and intervene with adolescents' abuse of alcohol and other chemical substances. The first step is taking an effective alcohol and drug-use history.

Treatment of levator syndrome using high-voltage electrogalvanic stimulation

Twenty patients diagnosed as having levator syndrome were treated with electrogalvanic stimulation (EGS). Many of these patients had previously been treated unsuccessfully with other methods. While 60 percent of patients had good or excellent response to EGS when assessed immediately after therapy, a subsequent telephone survey revealed that one third of this group experienced recurrence. Of the group of 20 patients, only five (25 percent) remained symptom-free. Though results in this study were less favorable than other reports, EGS may still serve as a valuable adjunct in the treatment of levator syndrome.

Acquired renal cystic disease in renal transplant recipients

End-stage kidneys are known to undergo cystic transformation in patients treated with long-term hemodialysis. Little has been published, however, on the occurrence of acquired renal cystic disease (ACD) in renal transplant recipients. The available clinical and histopathological data were gathered on 22 renal transplant recipients who had undergone either necropsy or nephrectomy. None of our patients exhibited cystic changes in their donor kidneys. However, 11 patients (50%) had ACD of the native kidneys while the remaining patients lacked ACD. The duration of dialysis prior to transplantation was significantly longer and the life of the functional transplant kidney was significantly shorter in the ACD group as compared to the noncystic group. The combined duration of ESRD (dialysis + transplantation), however, was comparable in the two groups. It thus appears that the presence of a functioning renal allograft somehow retards the evolution of cystic changes in the diseased native kidneys. 1 of the patients in the ACD group exhibited superimposed multifocal clear cell carcinoma of the affected kidney, while none of the patients in the noncystic group exhibited renal neoplasm.

Malignancy in uremia: dialysis versus transplantation

The incidence of cancer was compared in 499 dialysis patients and 121 renal transplant recipients. De novo malignancy developed in 15 patients on chronic dialysis (3 per cent) and in 6 transplant recipients (4.9 per cent), a significant increase over the expected number in the age-matched general population. There was no difference in the incidence of cancer in uremic patients on dialysis or after transplantation. A total of 10 dialysis patients (67 per cent) and 1 transplant patient (16 per cent) died of cancer. Neoplasms in the dialysis patients were the common types of mesenchymal tumors, while superficial skin cancers were seen more frequently in the transplant recipients. The differences in tumor types accounted for the higher mortality rate from cancer in the dialysis patients and may reflect different patterns of immunosuppression in these 2 patient populations.