Evaluation of Three State-of-the-Art Classifiers for Recognition of Activities of Daily Living from Smart Home Ambient Data

Smart homes for the aging population have recently started attracting the attention of the research community. The "health state" of smart homes is comprised of many different levels; starting with the physical health of citizens, it also includes longer-term health norms and outcomes, as well as the arena of positive behavior changes. One of the problems of interest is to monitor the activities of daily living (ADL) of the elderly, aiming at their protection and well-being. For this purpose, we installed passive infrared (PIR) sensors to detect motion in a specific area inside a smart apartment and used them to collect a set of ADL. In a novel approach, we describe a technology that allows the ground truth collected in one smart home to train activity recognition systems for other smart homes. We asked the users to label all instances of all ADL only once and subsequently applied data mining techniques to cluster in-home sensor firings. Each cluster would therefore represent the instances of the same activity. Once the clusters were associated to their corresponding activities, our system was able to recognize future activities. To improve the activity recognition accuracy, our system preprocessed raw sensor data by identifying overlapping activities. To evaluate the recognition performance from a 200-day dataset, we implemented three different active learning classification algorithms and compared their performance: naive Bayesian (NB), support vector machine (SVM) and random forest (RF). Based on our results, the RF classifier recognized activities with an average specificity of 96.53%, a sensitivity of 68.49%, a precision of 74.41% and an F-measure of 71.33%, outperforming both the NB and SVM classifiers. Further clustering markedly improved the results of the RF classifier. An activity recognition system based on PIR sensors in conjunction with a clustering classification approach was able to detect ADL from datasets collected from different homes. Thus, our PIR-based smart home technology could improve care and provide valuable information to better understand the functioning of our societies, as well as to inform both individual and collective action in a smart city scenario.

Intersystem crossing rates of S1 state keto-amino cytosine at low excess energy

The amino-keto tautomer of supersonic jet-cooled cytosine undergoes intersystem crossing (ISC) from the v = 0 and low-lying vibronic levels of its S1(¹ππ*) state. We investigate these ISC rates experimentally and theoretically as a function of S1 state vibrational excess energy Eexc. The S1 vibronic levels are pumped with a ~5 ns UV laser, the S1 and triplet state ion signals are separated by prompt or delayed ionization with a second UV laser pulse. After correcting the raw ISC yields for the relative S1 and T1ionization cross sections, we obtain energy dependent ISC quantum yields Q corr ISC =1% –5%. These are combined with previously measured vibronic state-specific decay rates, giving ISC rates kISC = 0.4–1.5 ⋅ 10⁹ s⁻¹, the corresponding S1⇝S0internal conversion (IC) rates are 30–100 times larger. Theoretical ISC rates are computed using SCS-CC2 methods, which predict rapid ISC from the S1; v = 0 state with kISC = 3 ⋅ 10⁹ s⁻¹ to the T1(³ππ*) triplet state. The surprisingly high rate of this El Sayed-forbidden transition is caused by a substantial admixture of ¹nOπ* character into the S1(¹ππ*) wave function at its non-planar minimum geometry. The combination of experiment and theory implies that (1) below Eexc = 550 cm⁻¹ in the S1 state, S1⇝S0internal conversion dominates the nonradiative decay with kIC ≥ 2 ⋅ 10¹⁰ s⁻¹, (2) the calculated S1⇝T1 (¹ππ*⇝³ππ*) ISC rate is in good agreement with experiment, (3) being El-Sayed forbidden, the S1⇝T1 ISC is moderately fast (kISC = 3 ⋅ 10⁹ s⁻¹), and not ultrafast, as claimed by other calculations, and (4) at Eexc ~ 550 cm⁻¹ the IC rate increases by ~50 times, probably by accessing the lowest conical intersection (the C5-twist CI) and thereby effectively switching off the ISC decay channels.

Adolescent Meaning Making of Past Experience

This study examined the meaning making processes of self-defining memories in adolescents, as well as how they co-construct the narratives of these events with their parents. The sample consisted of 53 students, aged 12-14, who came in for recorded laboratory sessions to discuss self-defining memories with their parents. These sessions were later coded on levels of meaning making and co-construction. These codes were, then, analyzed with the adolescents’ questionnaire scores regarding friendship quality, internalizing, and externalizing behaviors. The data revealed that adolescents and parents were both rated higher for more complex levels of meaning making and that those rated higher for more complex meaning making abilities had better friendship qualities. The implications of these findings were discussed in terms of their importance for parents supporting their children’s emotional expressivity, narrative abilities, and meaning making strategies.

The Making of Optimal and Consistent Policy: An Analytical Framework for Monetary Models

This paper shows that optimal policy and consistent policy outcomes require the use of control-theory and game-theory solution techniques. While optimal policy and consistent policy often produce different outcomes even in a one-period model, we analyze consistent policy and its outcome in a simple model, finding that the cause of the inconsistency with optimal policy traces to inconsistent targets in the social loss function. As a result, the central bank should adopt a loss function that differs from the social loss function. Carefully designing the central bank s loss function with consistent targets can harmonize optimal and consistent policy. This desirable result emerges from two observations. First, the social loss function reflects a normative process that does not necessarily prove consistent with the structure of the microeconomy. Thus, the social loss function cannot serve as a direct loss function for the central bank. Second, an optimal loss function for the central bank must depend on the structure of that microeconomy. In addition, this paper shows that control theory provides a benchmark for institution design in a game-theoretical framework.

The Making of Optimal and Consistent Policy: An Implementation Theory Framework for Monetary Policy

This paper shows that optimal policy and consistent policy outcomes require the use of control-theory and game-theory solution techniques. While optimal policy and consistent policy often produce different outcomes even in a one-period model, we analyze consistent policy and its outcome in a simple model, finding that the cause of the inconsistency with optimal policy traces to inconsistent targets in the social loss function. As a result, the social loss function cannot serve as a direct loss function for the central bank. Accordingly, we employ implementation theory to design a central bank loss function (mechanism design) with consistent targets, while the social loss function serves as a social welfare criterion. That is, with the correct mechanism design for the central bank loss function, optimal policy and consistent policy become identical. In other words, optimal policy proves implementable (consistent).

Sergei Rachmaninoff's Portraits from the collection of the State Central Museum of Musical Culture (Glinka Museum)

Signatur des Originals: S 36/F12277

Sergei Rachmaninoff's Portraits from the collection of the State Central Museum of Musical Culture (Glinka Museum)

Signatur des Originals: S 36/F12278