Strong minimax lower bounds for learning

Minimax lower bounds for concept learning state, for example, thatfor each sample size $n$ and learning rule $g_n$, there exists a distributionof the observation $X$ and a concept $C$ to be learnt such that the expectederror of $g_n$ is at least a constant times $V/n$, where $V$ is the VC dimensionof the concept class. However, these bounds do not tell anything about therate of decrease of the error for a {\sl fixed} distribution--concept pair.\\In this paper we investigate minimax lower bounds in such a--stronger--sense.We show that for several natural $k$--parameter concept classes, includingthe class of linear halfspaces, the class of balls, the class of polyhedrawith a certain number of faces, and a class of neural networks, for any{\sl sequence} of learning rules $\{g_n\}$, there exists a fixed distributionof $X$ and a fixed concept $C$ such that the expected error is larger thana constant times $k/n$ for {\sl infinitely many n}. We also obtain suchstrong minimax lower bounds for the tail distribution of the probabilityof error, which extend the corresponding minimax lower bounds.

Minimax lower bounds for the two-armed bandit problem

We obtain minimax lower bounds on the regret for the classicaltwo--armed bandit problem. We provide a finite--sample minimax version of the well--known log $n$ asymptotic lower bound of Lai and Robbins. Also, in contrast to the log $n$ asymptotic results on the regret, we show that the minimax regret is achieved by mere random guessing under fairly mild conditions on the set of allowable configurations of the two arms. That is, we show that for {\sl every} allocation rule and for {\sl every} $n$, there is a configuration such that the regret at time $n$ is at least 1 -- $\epsilon$ times the regret of random guessing, where $\epsilon$ is any small positive constant.

Cooperative Lakes Area Monitoring Project-Lower Gar Lake, 2007

Water fact sheet for Iowa Department of Natural Resources and the Geological Bureau.

Variscan lamprophyres in the Lower Penninic domain (Central Alps): age and tectonic significance

Lamprophyre dykes have been recently discovered in blocks of gneiss embedded in a calcschist formation of wildflysch type that forms the top of the Mesozoic-Tertiary metasedimentary cover of the Antigorio nappe (the Teggiolo zone) in the Val Bavona (Lower Penninic, NW Ticino, Switzerland). The presence of the lamprophyres gives a clue to the possible source of these blocks. Similar dykes occur in the N part of the Maggia nappe where they are intruded into the Matorello granite and the surrounding gneisses. We studied these lamprophyres at two localities in the Teggiolo zone (Tamierpass and Lago del Zott) and at one locality in the Maggia nappe (Laghetti). Detailed mineralogical and geochemical investigations confirm their great similarity, particularly between the Tamier and Laghetti dykes. They all recrystallized during Alpine metamorphism under amphibolite facies conditions and lost their primary mineral assemblages and textures. The chemistry reveals a calc-alkaline affinity, a limited differentiation range, features of mineral accumulation and intense remobilization in some cases. The lamprophyres are characterized by a high mg# and relatively low contents in REE and other incompatible elements. In situ SHRIMP and LA-ICPMS U-Pb zircon dating yielded ages of 284.8 +/- 1.7 Ma (Tamier), 290.0 +/- 1.3 Ma (Zott) and 290.5 +/- 3.7 Ma (Laghetti). These ages are compatible with the general late- to post-Variscan magmatic evolution of the Helvetic and Lower Penninic domains. The lamprophyres are considered as melts derived from the lithospheric Variscan mantle, variously hybridized and differentiated at the contact with crustal material during late- to post-orogenic extension. These lamprophyres are chemically distinct from earlier lamprophyres of Visean age, emplaced together with their associated granites in transcurrent fault zones during the Variscan orogenic compression. The similarity of these different dykes suggests that the front of the Maggia nappe is a likely source of the gneissic blocks embedded in the calcschists at the top of the Teggiolo zone. They would have been provided by the advancing Maggia nappe during its thrusting over the Antigorio nappe and simultaneous closure of the Teggiolo sedimentary basin.

Step count is associated with lower nighttime systolic blood pressure and increased dipping.

BACKGROUND: Higher nighttime blood pressure (BP) and the loss of nocturnal dipping of BP are associated with an increased risk for cardiovascular events. However, the determinants of the loss of nocturnal BP dipping are only beginning to be understood. We investigated whether different indicators of physical activity were associated with the loss of nocturnal dipping of BP. METHODS: We conducted a cross-sectional study of 103 patients referred for 24-hour ambulatory monitoring of BP. We measured these patients' step count (SC), active energy expenditure (AEE), and total energy expenditure simultaneously, using actigraphs. RESULTS: In our study population of 103 patients, most of whom were hypertensive, SC and AEE were associated with nighttime systolic BP in univariate (SC, r = -0.28, P < 0.01; AEE, r = -0.20, P = 0.046) and multivariate linear regression analyses (SC, coefficient beta = -5.37, P < 0.001; AEE, coefficient beta = -0.24, P < 0.01). Step count was associated with both systolic (r = 0.23, P = 0.018) and diastolic (r = 0.20, P = 0.045) BP dipping. Nighttime systolic BP decreased progressively across the categories of sedentary, moderately active, and active participants (125mm Hg, 116mm Hg, 112mm Hg, respectively; P = 0.002). The degree of BP dipping of BP increased progressively across the same three categories of activity (respectively 8.9%, 14.6%, and 18.6%, P = 0.002, for systolic BP and respectively 12.8%, 18.1%, and 22.2%, P = 0.006, for diastolic BP). CONCLUSIONS: Step count is continuously associated with nighttime systolic BP and with the degree of BP dipping independently of 24-hour mean BP. The combined use of an actigraph for measuring indicators of physical activity and a device for 24-hour measurement of ambulatory BP may help identify patients at increased risk for cardiovascular events in whom increased physical activity toward higher target levels may be recommended.

Help For Lower Income Iowans With Medicare, 2007

If you have limited income and resources, the State of Iowa may pay some of your Medicare expenses. Programs paying these expenses are shown in the chart on page 2. To see if you might be eligible, answer the questions below.

Groundwater Availability Modeling: Lower Dakota Aquifer, 2008

This monthly report from the Iowa Department of Natural Resources is about the water quality management of Iowa's rivers, streams and lakes.

Which clinical signs predict hypoxaemia in young Senegalese children with acute lower respiratory tract disease?

BACKGROUND: Acute lower respiratory tract diseases are an important cause of mortality in children in resource-limited settings. In the absence of pulse oximetry, clinicians rely on clinical signs to detect hypoxaemia. OBJECTIVE: To assess the diagnostic value of clinical signs of hypoxaemia in children aged 2 months to 5 years with acute lower respiratory tract disease. METHODS: Seventy children with a history of cough and signs of respiratory distress were enrolled. Three experienced physicians recorded clinical signs and oxygen saturation by pulse oximetry. Hypoxaemia was defined as oxygen saturation <90%. Clinical predictors of hypoxaemia were evaluated using adjusted diagnostic odds ratios (aDOR). RESULTS: There was a 43% prevalence of hypoxaemia. An initial visual impression of poor general status [aDOR 20·0, 95% CI 3·8-106], severe chest-indrawing (aDOR 9·8, 95% CI 1·5-65), audible grunting (aDOR 6·9, 95% CI 1·4-25) and cyanosis (aDOR 26·5, 95% CI 1·1-677) were significant predictors of hypoxaemia. CONCLUSION: In children under 5 years of age, several simple clinical signs are reliable predictors of hypoxaemia. These should be included in diagnostic guidelines.

Physical activity in daily life is associated with lower adiposity values than doing weekly sports in Lc65+ cohort at baseline.

BACKGROUND: Overweight and obesity prevalence is the highest at age 65-75 years in Lausanne (compared with younger classes). We aimed to describe 1) eating habits, daily physical activity (PA), and sports frequency in community-dwelling adults aged 65-70, 2) the links of these behaviors with socio-economic factors, and 3) with adiposity. METHODS: Cross-sectional analysis of Lc65+ cohort at baseline, including 1260 adults from the general population of Lausanne aged 65-70 years. Eating habits (8 items from MNA) and PA (sports frequency and daily PA: walking and using stairs) were assessed by questionnaires. Body mass index (BMI), supra-iliac (SISF), triceps skin-folds (TSF), waist circumference (WC), and WHR were measured. RESULTS: Prevalence of overweight (BMI 25.0-29.9 kg/m2), obesity (BMI ≥ 30.0 kg/m2), and abdominal obesity was 53%, 24%, and 45% in men; 35%, 23%, and 45% in women.Intake of fruits or vegetables (FV) ≥ twice/day was negatively associated with male sex (prevalence 81% versus 90%, chi-square P < 0.001). The proportion avoiding stairs in daily life was higher among women (25%) than among men (20%, chi-square P=0.003).In multivariate analyses among both sexes, eating FV, using stairs in daily life ("stairs"), and doing sports ≥ once/week were significantly negatively associated with financial difficulties (stairs: OR=0.54, 95% CI=0.40-0.72) and positively with educational level (stairs: OR=1.68, 95% CI=1.17-2.43 for high school).For all five log-transformed adiposity indicators in women, and for all indicators except SISF and TSF in men, a gradual decrease in adiposity was observed from category "no stairs, sports < once/week" (reference), to "no stairs, sports ≥ once/week", to "stairs, sports < once/week", and "stairs, sports ≥ once/week" (for example: WC in men, respectively: ß= -0.03, 95% CI= -0.07-0.02; ß= -0.06, 95% CI= -0.09- -0.03; ß= -0.10, 95% CI= -0.12- -0.07). CONCLUSIONS: In this population with high overweight and obesity prevalence, eating FV and PA were strongly negatively associated with financial difficulties and positively with education. Using stairs in daily life was more strongly negatively associated with adiposity than doing sports ≥ once/week.

Obese and fit adolescents have lower blood pressure levels than obese and unfit counterparts.

The aim of the study was to assess the effects of physical fitness on the relationships between body mass index (BMI) and body fat (BF) on blood pressure (BP) levels. Cross-sectional study conducted in 25 schools of Lisbon (Portugal), including 2041 boys and 1995 girls aged 10-18. BF was assessed by bioimpedance. Cardiovascular fitness was assessed by the 20-meter shuttle run and classified as fit/unfit. Obesity (BMI or BF defined) was defined according to international criteria. In both sexes, BMI was positively related with systolic and diastolic BP, while BF was only positively related with diastolic BP z-scores. No interaction was found between fitness and BMI categories regarding BP levels, while for BF a significant interaction was found. Being fit reduced the BF-induced increase in the Odds ratio (OR) of presenting with high BP: OR (95% confidence interval) 1.01 (0.73-1.40) and 0.99 (0.70-1.38) for overweight and obese fit boys, respectively, the corresponding values for unfit overweight and obese boys being 1.33 (0.94-1.90) and 1.75 (1.34-2.28), respectively. The values were 0.88 (0.57-1.35) and 1.66 (0.98-2.80) for overweight and obese fit girls, respectively, the corresponding values for unfit overweight and obese being 1.63 (1.12-2.37) and 1.90 (1.32-2.73) respectively. No interaction was found between fitness and BMI-defined overweight and obesity. Being fit reduces the negative impact of BF on BP levels and high BP status in adolescents. This protective effect was not found with BMI.