Efficient search for statistically significant dependency rules in binary data

Analyzing statistical dependencies is a fundamental problem in all empirical science. Dependencies help us understand causes and effects, create new scientific theories, and invent cures to problems. Nowadays, large amounts of data is available, but efficient computational tools for analyzing the data are missing. In this research, we develop efficient algorithms for a commonly occurring search problem - searching for the statistically most significant dependency rules in binary data. We consider dependency rules of the form X->A or X->not A, where X is a set of positive-valued attributes and A is a single attribute. Such rules describe which factors either increase or decrease the probability of the consequent A. A classical example are genetic and environmental factors, which can either cause or prevent a disease. The emphasis in this research is that the discovered dependencies should be genuine - i.e. they should also hold in future data. This is an important distinction from the traditional association rules, which - in spite of their name and a similar appearance to dependency rules - do not necessarily represent statistical dependencies at all or represent only spurious connections, which occur by chance. Therefore, the principal objective is to search for the rules with statistical significance measures. Another important objective is to search for only non-redundant rules, which express the real causes of dependence, without any occasional extra factors. The extra factors do not add any new information on the dependence, but can only blur it and make it less accurate in future data. The problem is computationally very demanding, because the number of all possible rules increases exponentially with the number of attributes. In addition, neither the statistical dependency nor the statistical significance are monotonic properties, which means that the traditional pruning techniques do not work. As a solution, we first derive the mathematical basis for pruning the search space with any well-behaving statistical significance measures. The mathematical theory is complemented by a new algorithmic invention, which enables an efficient search without any heuristic restrictions. The resulting algorithm can be used to search for both positive and negative dependencies with any commonly used statistical measures, like Fisher's exact test, the chi-squared measure, mutual information, and z scores. According to our experiments, the algorithm is well-scalable, especially with Fisher's exact test. It can easily handle even the densest data sets with 10000-20000 attributes. Still, the results are globally optimal, which is a remarkable improvement over the existing solutions. In practice, this means that the user does not have to worry whether the dependencies hold in future data or if the data still contains better, but undiscovered dependencies.

The new binary pulsar and the observation of gravitational spin precession

The possibility of observing gravitational spin precession due to spin-orbit coupling in a binary pulsar system is considered. An analysis is presented which can aid in delineating the relevant physical effects from pulse-structure data. In this analysis, it is assumed that the pulsar radiation emanates from a cone whose axis is tilted with respect to the axis of rotation. It is found that the time-averaged pulse width and polarization sweep vary periodically with time and that this variation has a periodicity of the order of the spin-precession frequency averaged over a complete revolution. It is concluded that for an orbital period of about 180 years, it suffices to measure polarization data with an accuracy of a few parts in 100 over a period of six months to a year in order to uncover the effects of spin precession. The consistency of the analysis is checked, and the calculations are applied to a recently discovered binary pulsar.

Quantization noise in binary holograms

The signal-to-noise (S/N) ratio in the reconstructed image from a binary hologram has been quantitatively related to the amplitude and phase quantization levels. The S/N ratio increases monotonically with increasing number of quantization levels. This observation is further supported by experimental results.

Critical Exponent β and Rectilinear Diameter of the Binary-Liquid System Carbon Disulfide + Nitromethane

The system CS2 + CH3NO2 shows β=0.315±0.004 over 10-6<ε=|T-Tc| / Tc<2�10-1 with no indication of a classical value ½ even far away from Tc. The diameter shows a curvature and is of the form �c+b ε+fε7 / 8exp(-gεh).

Dispersion of the stress-optic coefficient in glasses

The stress-optic coefficient (n3/2)(q11-q12) has been determined for a series of 18 optical glasses of different compositions in the wavelength range 5700-3200 Å. The coefficients are negative for all the glasses except for a high-lead-content glass of density 6·7 and refractive index 1·89. The numerical value of the coefficient decreases as one proceeds to the ultraviolet. This behaviour is just the opposite of what is observed in fused silica. By applying Mueller's theory, the strain polarizability constant and its dispersion have been evaluated.

Critical point phenomena in the electrical resistivity of binary liquid systems: Cs2 + CH3CN and CS2 + CH3NO2

Electrical resistance measurements are reported on the binary liquid mixtures CS2 + CH3CN and CS2 + CH3NO2 with special reference to the critical region. Impurity conduction seems to be the dominant mechanism for charge transport. For the liquid mixture filled at the critical composition, the resistance of the system aboveT c follows the relationR=R c−A(T−T c) b withb=0·6±0·1. BelowT c the conductivities of the two phases obey a relation σ2−σ1=B(T c−T)β with β=0·34±0·02, the exponent of the transport coefficient being the same as the exponent of the order parameter, an equilibrium property.

Logical design of a negative binary adder-subtracter

Two different designs for negative binary adder-subtracter are compared. Ono design uses the method of a hybrid-carry—borrow, while the other 11303 the method of polarization and addition.

Critical behaviour of electrical resistivity in 5 binary liquid systems

For five binary liquid systems CS2+CH3CN, CS2+CH3NO2, CS2+(CH3CO)2O, C6H12+(CH3CO)2O, n-C7H16+(CH3CO)2O, the electrical resistance has been measured near the critical solution temperatures. The behaviour is universal. Below Tc, the conductivities of the two phases follow σ1−σ2 β, where = T−Tc Tc with β≈0.35. In the one phase region with b≈0.35±0.1 and is positive in some cases and negative in others.

Scaling for binary liquid systems and ferromagnets under pressure

Scaling relations between the critical indices are derived for two similar systems exhibiting λ lines: binary liquid systems and ferromagnets under pressure. In addition to the usual scaling relations, this procedure gives information about other weakly divergent quantities like isothermal compressibility and thermal expansion. Suggestions for more detailed investigations are made.

Critical Exponent β and Rectilinear Diameter of the Binary-Liquid System Carbon Disulfide + Nitromethane

The system CS2 + CH3NO2 shows β=0.315±0.004 over 10-6<ε=|T-Tc| / Tc<2-10-1 with no indication of a classical value ½ even far away from Tc. The diameter shows a curvature and is of the form - c+b ε+fε7 / 8exp(-gεh).

Dielectric properties of Li2O-3B(2)O(3) glasses

The frequency and temperature dependences of the dielectric constant and the electrical conductivity of the transparent glasses in the composition Li2O-3B(2)O(3) were investigated in the 100 Hz-10 MHz frequency range. The dielectric constant and the loss in the low frequency regime were electrode material dependent. Dielectric and electrical relaxations were, respectively, analyzed using the Cole-Cole and electric modulus formalisms. The dielectric relaxation mechanism was discussed in the framework of electrode and charge carrier (hopping of the ions) related polarization using generalized Cole-Cole expression. The frequency dependent electrical conductivity was rationalized using Jonscher's power law. The activation energy associated with the dc conductivity was 0.80 +/- 0.02 eV, which was ascribed to the motion of Li+ ions in the glass matrix. The activation energy associated with dielectric relaxation was almost equal to that of the dc conductivity, indicating that the same species took part in both the processes. Temperature dependent behavior of the frequency exponent (n) suggested that the correlated barrier hopping model was the most apposite to rationalize the electrical transport phenomenon in Li2O-3B(2)O(3) glasses. These glasses on heating at 933 K/10 h resulted in the known nonlinear optical phase LiB3O5.

Study of As2(Se,Te)3 glasses by X-ray absorption and photoelectron spectroscopy

Arsenic selenide-telluride glasses have been investigated by X-ray absorption and photoelectron spectroscopy. The core electron energy shifts and chemical shifts in K-absorption edge measurements associated with the glass-crystal transitions of pure As2Se3 and As2Te3 have been studied. The effect of composition on the core level energy and valence bands of As2(Se,Te)3 glasses, has been discussed. Mixed-composition glasses are found to be considerably ionic.

Rapid quenching and glass formation in chalcogenide glasses: Preparation and properties of Ge-As-Te glasses over an extended composition ranges

In Ge-As-Te system, the glass forming region determined by normal melt quenching method has two regions (GFR I and GFR II) separated by few compositions gap. With a simple laboratory built twin roller apparatus, we have succeeded in preparing Ge7.5AsxTe92.5-x glasses over extended composition ranges. A distinct change in T-g is observed at x = 40, exactly at which the separation of the glass forming regions occur indicating the changes in the connectivity and the rigidity of the structural network. The maximum observed in glass transition (T-g) at x = 55 corresponding to the average coordination number (Z(av)) = 2.70 is an evidence for the shift of the rigidity percolation threshold (RPT) from Z(av) = 2.40 as predicted by the recent theories. The glass forming tendency (K-gl) and Delta T (=T-c-T-g) is low for the glasses in the GFR I and high for the glasses in the GFR II.

Evidence of an intermediate phase in ternary Ge7Se93-xSbx glasses

The compositional dependence of thermal properties, such as glass transition temperature (T-g), non-reversing enthalpy change (Delta H-NR) and the specific heat capacity change (Delta C-p) of melt quenched Ge7Se93-xSbx (21 a parts per thousand currency sign x a parts per thousand currency sign 31) glasses, has been studied using alternating differential scanning calorimetry (ADSC) which is analogous to modulated differential scanning calorimetry (MDSC). The glass transition temperature, T-g, which is a measure of global connectivity of the glass, has been found to increase with the addition of Sb. In addition, a change in slope has been observed in the composition dependence of T-g at an average coordination aOE (c) r > = 2.40. The experimentally observed compositional variation of glass transition temperature, has been compared with the theoretical predictions from the stochastic agglomeration theory (SAT) and has been found to be consistent. Further, a narrow thermally reversing window is seen in the compositional variation of the relaxation enthalpy (Delta H-NR), which is centered around aOE (c) r > = 2.40. The change in specific heat capacity (Delta C-p) at T-g is also found to exhibit a distinct minima at aOE (c) r > = 2.40, suggesting that the structural rearrangements for the liquid in the glass transition region are minimized around aOE (c) r > = 2.4.